Providing Self-Monitoring Feedback

 

Author’s Name:

 

Institution:

 

Month and Year of Submission:

 

Abstract

Mobile devices can be made use of in so many other ways apart from passing information from one person to another through call or text messages. The devices have been created in ways that can capture and interpret data of various forms ranging from letters, numeric, videos, and images. The devices can be used in scanning to read codes on various items in supermarkets with the help of configured applications. However, in the present world, most people are suffering from various diseases that are caused by the poor eating habits. Some of the eating habits leading to poor health and suffering from diseases are lack of balanced diets that contain the right nutrients. The research will examine the level of calories intake by a person on every meal. Certain people may take more or less of calories depending on the best meals every person is served. To capture and make the best use of use of mobile phone devices, a prototype app will be made to capture selected information and thereafter a picture of the specified meal taken. The results are calculated based on the inspection of the picture in conjunction with the offered information. A majority of the population in the world poses Android mobile phones that can be used to collect such information and offer feedback to the patient or user without visiting a specialist for advice. “(Word count as specified in the Course Handbook: 14112 words).”

 

Acknowledgment

This dissertation has made me make a great milestone in my academics. The knowledge and information I have gathered and learned have been influenced by my participation in the research of this project. I offer gratitude to everyone who played a part in ensuring that I have gained the required knowledge.

My thanks go to my advisor, who advises me on the course of my research. The advisor has been available whenever I needed him during the research proposal preparation and during the research’s conceptualization. If it were not the efforts applied by my advisor, I would not have made to accomplish what I have achieved. The instructions, guidance, and recommendations offered enabled me to gather the right information in the preparation of my dissertation.

My gratitude also goes to my instructors, professors, and tutors who took part in my journey towards the accomplishment of my course. They all believed in my ability to make it through my studies and I now thank them all for the motivational guidance and encouragement. Finally, I thank my family, guardians, and my sponsors for facilitating my education to the level I am. They contributed as much as they could in terms of motivation and finances to see me through the system and achieve the best grade. I am so grateful, thank you all, and be blessed.

 

Contents

Abstract ii

Acknowledgment iii

List of Figures. vi

List of Tables. vii

Chapter One. 1

1.0 Introduction. 1

1.1 Background Information. 1

1.2 The Importance of Lightweight Visual Data Analysis on Mobile Devices. 2

1.3 The Theoretical basis for Visual Data Analysis. 3

1.4 Visual Data Analysis Statement. 5

1.6 The Relevance and Effectiveness of Lightweight Visual Data Analysis Study on Mobile Devices  6

1.7 Hypothesis and Research Questions. 6

1.7.1 Hypothesis. 6

1.7.2 Research Questions. 7

Chapter Two. 9

2.0 Literature Review.. 9

2.1 Background. 9

2.3 Theories on Visual Data Analysis. 10

2.3.1 Data Analysis Exploration. 10

2.3.2 Self-Monitoring Visual Data Analytics. 11

2.3.3 Monitoring Health using Analyzed Visual Data. 13

2.4 Data Exploration and Analysis Literature. 16

2.5 Visual Data Extraction for Analysis. 17

2.6 Visual Data Scalability Analysis Review.. 19

Chapter Three. 22

3.0 Methodology. 22

3.1 Visual Data Analysis Tools. 22

3.1.1 Tableau. 22

3.1.2 Data-Driven Documents (D3.js). 22

3.1.3 WebDataRocks. 24

3.1.4 BIRT.. 25

3.1.5 Google Charts. 25

3.1.6 Cytoscape.js. 25

3.2 Mobile Device Visualization Approaches. 26

3.2.1 Approaches of Compact Visualizations. 27

3.3 Cutting-edge Visualizations in lieu of Mobile Interaction. 28

3.4 Dietary Intake and Its Contribution to Weight Loss through Self-monitoring. 29

3.5 The range of Self-monitoring mobile device application. 30

3.6 An Investigation of the Effect of Mobile Devices on Self-monitoring. 31

3.7 Self-monitoring and Visual Data Analysis of Dietary Intake. 31

3.8 Self-monitoring and Visual Data Analysis of Physical Activity. 32

3.9 Self-monitoring and Visual Data Analysis of Weight. 32

Chapter Four. 34

4.0 Data Analysis and Discussion. 34

4.1 Quantitative Visual Data Analysis Using Regression Models. 34

4.2 Surveys on Human Effect Caused by Handheld Mobile Devices. 38

4.3 Discussion. 40

Chapter Five. 44

5.0 Conclusions and Further Research. 44

5.1 Conclusion. 44

5.2 Future Research. 45

References. 48

Bibliography. 51

Appendices. 52

 

List of Figures

Figure 3. 1 Bubble Chart. 23

Figure 3. 2 Scatterplot Chart. 23

Figure 4. 1 Graphical Representation of Results. 37

Figure 4. 2 Calories Counter. 37

Figure 4. 3 Quantity of hours spent on hand held devices (HHDs) 39

List of Tables

Table 4. 1 D3.JS vs. TABLEAU_ 24

Table 4. 2 Sample Population dataset_ 35

Table 4. 3 Descriptive Statistics_ 36

Table 4. 4 Summarized Demographic Data_ 38

Table 4. 5 Purpose of mobile device_ 39

Table 4. 6 Discomforts encountered_ 40

Table 4. 7 Tangling sensation_ 40

 

1.1 Background Information

Health is a major concern that everyone in the society must take part in making it better. Starting with the nutrients of meals, the location where meals take place, and eating habits and mechanisms must be analyzed be analyzed using a developed app or database. The people need to be informed on the nutrients and other components found in food. People have the ability of taking or collecting visual data using mobile devices but nothing much has been done regarding the visual data analysis in form of generating feedback to users. The concept of getting recommendations and analysis reports from captured data in form of images can aid the larger population in saving time and other resources. Certain items like those that set dining tables and plates served with meals have always been a challenge in detecting the nutrients. Mobile phones and other wireless machines with sensors have the ability to scan and read hidden information through codes and other specified credentials as chosen by the users. Mobile phones have the same ability to make use of pictures and data to help calculate the level of calories consumed by a person on the meals served. A meal with various ingredients will be expected to have certain levels of calories based on the standard recipes for preparing such meals. For example, a person who has been served breakfast containing an egg, milk, and bread can be informed of the amount of calorie consumed. A lightweight data capture and visualization project will be done to gather information through text and pictures to analyze the number of calories consumed. Despite the pictures having various challenges such as darkness, too much light, or blurred, the information collected is analyzed hand in hand with the selected items purported to be in the pictured meal. The design of the app has no specific details to contain but the ability to analyze data using some of the keyed information related to taken pictures. Certain chronic diseases in the present age are caused by poor dieting and lack of exercise to burn the extra calories taken. Instead of keeping records on the served meal to inform nutritionists, individual people can have real-time tracking of the meals taken or even before eating to ensure the right quantities of calories have been served to avoid gaining weight and becoming obese. Overweight is a major cause of the major diseases affecting the world like high blood pressure, diabetes, and to some levels the causes of severe cancers. Every Android phone user has the ability to read and respond to various text information displayed on the screen of the phone. Making the app simple to collect a few related or present items found on the served meal then taking a picture will help to give an idea on the number of calories consumed.

1.2 The Importance of Lightweight Visual Data Analysis on Mobile Devices

We have a variety of foods eaten in the world that cannot be analyzed from one central place. Assessing on the nutrient components has been a major challenge. With a well-trained artificial intelligence program, certain nutrients can be read and analyzed from visuals taken in form of images or pictures. Due to the extended use of mobile devices to a great population in the world, performing analysis-using pictures can be aided with the available devices. Almost every person within a set diner or served meal owns a mobile device capable of collecting visual data. Instant responses and feedback offer efficient and reliable information that can be trusted than getting results from samples dropped in a laboratory. The generated results on an immediate effect can lead to trustworthiness and reliability of the self-monitored feedback. The study will be of significance to all mobile devices users because they will make good use of the gadgets in assessing the quality and value of meals consumed. Instead of taking pictures of junk and fast foods to post on social media, educative information can be generated from the pictures and guide towards the right and required diet intake. Upon understanding the number of calories contained in meals taken by a person every day, setting targets on what to feed on to regulate or balance the number of calories will lead to a healthy world.

The world is driven under the influence of technology and mobile devices have been including the abilities to capture information and processing according to the set commands. For example, data collections with applications that can recognize visuals make the exercise of learning and doing research easier and user-friendly. Getting trends from the set format and acceptable influxes create the desire of testing the available information.

The social sites have enabled people to pose and tell the world the meals they take but the ability to calculate and inform about nutrients content has been a challenge. For example, Instagram users have the habit of posting pictures and events attended. The meals served are taken pictures and posted on social media. Getting an application that can read, interpret, and analyze the information on such pictures can help such users in understanding the pictures and getting recommendations on what to do in making it right (Varona-Marin, Scott, & University of Waterloo. 2016).

The relationships between certain components on images have subtle correlation trends that must be understood by users of such information. Besides the decorations and fancy presentations of visuals, performing a visualization of data analysis helps to understand the deep meaning of how human eyes can view objects of things. For example, if an application can have the ability to discover an ingredient used on a meal and calculates the calorie content; such a system can be applied by the health ministries in ensuring that set standards of serving food are met in the hospitality industry.

The world does not want to waste time making research on certain topics such as maintaining healthy eating because the required efforts and resources cannot be met by everyone. Since technology has enabled the embedding of health and fitness applications on mobile devices, the people can feel at ease making use of applications that require a few minutes of their time even before engaging in an activity to take pictures and upload them into the application.

The information or feedback received from a visualized data analysis using mobile devices can be shared with other users as they can take screenshots or download results. Through the simple exercise of passing the results obtained to another person creates a platform for learning and getting more insights about topics affecting a majority of the population. The effect of having many calories in the body can facilitate weight gain and in some cases can lead to obesity (Varona-Marin, Scott, & University of Waterloo. 2016).

1.3 The Theoretical basis for Visual Data Analysis

Visual data analysis tools come with interfaces used to interact with users. The interface offers a screen to input data and takes pictures that can be processed and analyzed for the intended results. The screen offers choices to make while inserting the required information into an application. Depending on the complexity of the system developed, the input data or information may range from charts, line graphs, bar graphs, and feedback recommendations. The solutions offered by a visual data analysis application may include but not limited to the following:

No or little coding during the preparation and implementation of the analysis
The complexity of the analysis must be reduced and made simple to locate data from various sources
The graphics used must be easily customized and user attractive to encourage regular use of such tools in data analysis
All levels of underlying data must be drilled down with ease to enable the ability to receive detailed results
Multiple views must be combined in a possible manner to create an at-a-glance general understanding of the entire process used in extracting, processing, analyzing, and interpreting results.

Human beings have huge volumes of data lacking understanding until devices and applications or systems are made to solve the problem. For example, mobile devices have the ability to be installed apps that can perform much of the data collection and analysis based on the set commands and instructions used while creating the databases. Creating an interactive dashboard where collection and analysis of information can take place must be created. The information at hand can never have meaning without processing and generating results that can be understood by everyone in the simplest meaning. Meals taken on a daily basis can be easy to consume but difficult to determine the goodness or badness in them after taking. A visualization is a helpful tool because it helps users of such tools in making the best use of the visualized items in life into account while gaining benefits. For example, the collection of calories information from served meals before taking them enables people to learn the number of calories in every meal hence measuring the intake of nutrients becomes a hobby. The results obtained are received and an application of common sense used to determine the remedy or corrective measure to the issued feedback. For example, when a system response informs that the consumed ingredient is in less quantity than the recommended, the suggestive measure is to ensure more is consumed to meet the required standard.

1.4 Visual Data Analysis Statement

Health organizations in the world have determined the use of mobile devices in determining the health status of people. Since a majority of the population own mobile devices, the collection of information from meals can be easier because the required details are limited to a few options and taking snapshots. The images and information provided can offer insights on the nutrients contained in every meal serving and recommend for a reduction or increase in the consumption of various portions. An average recommendation of calories intake for a normal woman is 2000 calories and for a man 2500 calories in a day. The prototype application is expected to generate the results of the number of calories taken in a day. When the results are below the recommended average consumption, the application advises for an increase while when the results are above the recommended average, a reduction in consumption is advised as the feedback.
1.5 Lightweight Visual Data Analysis on Mobile Devices Study Objectives

The population of the people is increasing at a high rate and healthcare providers are on the run doing research on emerging diseases. Technology has become a basic knowledge tool where anyone can make use of programmed systems by submitting the requested information and the interpretation of results made easier to understand. Making physical visits to professionals to help in the determination of nutrients and calories intake by people on daily meals can be a difficult exercise, but using handheld devices owned by the person to analyze himself or herself could save time and make the process faster. The study will examine the impact of implementing a personal visual data analysis used on mobile devices. The study objective is to evaluate how people make use of mobile devices with visualized data. The images and other visuals captured can be utilized to process information and offer detailed results. The target population is all users of mobile devices. A great percentage of the people in the world own Android or Smartphones capable of collecting visual data information. The allocated memory size can be installed applications ready to analyze captured visuals of meals and calculate the calories content on the meals served. The visuals can be analyzed based on shape, color, and pixels of pictures taken.

1.6 The Relevance and Effectiveness of Lightweight Visual Data Analysis Study on Mobile Devices

The effectiveness of the visualizations used by lightweight mobile devices has not been fully discovered but with the current research, some sort of analysis can be done to establish and discover the contents in a meal. Due to the lack of interventions on the visualizations, an assist of a few input data is required to facilitate the comparison and analysis of the items within a meal. Making use of data analysis concepts, collection and processing will be the major tasks of the lightweight visual data analysis app will have a primary objective of collecting information in form of pictures assisted with selected text predictions to generate the number of calories in food. The data processing criteria have been integrated with the information about the number of calories contained in different kinds of food. An analytical process takes the development of data from input records or historical data. Considering the reasons why many people purchase phones are due to communication and other luxury options like taking pictures and listening to music. Making simple and less time-consuming applications that can benefit users regarding health and other real-life aspects can enable knowledge enhancement towards enlightening people. Making such an application to require less memory space and generate results run through a programmed application can make work easier. It can take so much time and effort to develop such an application but once accepted in the market and running the target organizations, institutions, or the market will help in generating revenue to maintain and make updates.

1.7 Hypothesis and Research Questions
1.7.1 Hypothesis

Life challenges can be solved by mathematical techniques to derive comparisons from two datasets. Under the approach of using visualizations to determine the calories taken in by people from the served meals, the techniques used can be applied on the involved variables such as the number of humans and the type of food. Under the visualization of human nature, the aspects directly influencing the people have a great influence on the determination of steps taken in life. The sample size population and the type of food served comprise the variables needed to determine the expected outcome of the visualized data. The tests available for visualized data analysis on mobile devices include the relation, determination of the amount of certain nutrients intake, and other aspects of life like the kind of exercise engaged in keeping fit. Using the American Heritage Dictionary in making the definition of a hypothesis, “a hypothesis is a tentative explanation for an observation, phenomenon, or scientific problem that can be tested by further investigation.” Due to the ever-changing world with the new approaches in technology, health matters have posed a major challenge because of the development of fast foods served on the streets and institutions. The teenagers are at a greater risk of getting involved in certain behavioral traits that lead to surviving on junk food. The hypothesis in our case is the determination of the number of calories on the meals served at different times of the day compared to the recommended daily average consumption. The objective of the study is to determine the number of calories consumed by people while taking meals. It has been determined that the high or low number of calories intake is a result of the lack of knowledge on the number of calories contained in every meal. Creating a system or application that can work on mobile devices will enable people to discover the number and learn better ways of managing the calories intake. Apart from the number of calories taken every day, burning the bad calories can be a challenge due to the lack of an available incorporated tool for checking the number of calories intake and the number of calories burnt.

1.7.2 Research Questions

The research on lightweight visual data analysis on mobile devices will be evaluating the theoretical supporting information concerning lightweight visual data analysis tools on mobile devices. It will be targeting to get possible answers to the following questions:

Does the use of mobile devices aid in visual data collection?
Do handheld devices have the ability to process visual data and generate easy to understand results?
Will the use of lightweight visual data analysis tools on mobile devices make the understanding of healthy eating habits and create a routine to check the contents in meals before choosing what to eat?
Will the generated results offer satisfactory feedback to mobile devices users?

The term visualize has two distinct meanings. The first meaning being the formation of object images mentally in an internal and cognitive manner while the second meaning occurs when the eyes have the ability to make things visible through the concept of an external perpetual role. The two definitions of visualized images can be assumed to take the relationship between cognition and perception; it can be argued that the visualization of images keeps changing with time. The use of graphical demonstrations of data and concepts are being used to generate results from seen images. Similarly, the objectives of visualization have experienced tremendous changes just like the meaning of the term. The goals of visualization are limited to three as follows:

Offering an exploratory analysis in a typically undirected search for the latest information from trends and structures without basing on any initial hypothesis
Making of a confirmatory analysis in the description of the existing hypothesis by making goal-oriented examinations with the intention of rejecting or making confirmations
Making presentations through an effective and efficient communication of ideas and facts fixed on a priori

The early uses of visuals like maps can date back the historic periods of various events, the major goal of offering such maps was to create a presentation of the past into the present. As much as presentations can offer accurate information, the clarity of images wears out with time and the existing knowledge fails to deliver the original message. Technology and computers brought about the rise of exploratory data analysis using graphical user interfaces. The statistical research was developed and made easier by the use of statistical tools developed and run on computers. Own research in the visualization of information started in the past two decades where the discipline was developed by individual people with the intention of making use of images to generate information. The search for interesting data with structures on various scenarios facilitated the move to performing the visual data analysis. The interaction of many structures with the mindset of what-if certain objects could have been seen in different ways has facilitated much in the search and development of visualized data analysis. Despite the search going on, certain interactions on the subject are in progress such as the linking of several visualizations, modifications of the run-time parameters, and data filtration to attain a standard form of incorporating such information in a customized manner.

2.3.1 Data Analysis Exploration

Visual data analytics is a science of reasoning that is facilitated by an interaction of visuals. Information gathering, processing, presentation, and making decisions are all involved in visual data analysis. The ultimate objective in visual data analysis is to ensure that the visible images have produced results on the intended expectations. For example, human health is highly determined by the diet taken. When images of dishes are taken and sent to a server for analysis and processing, the nutrient contents must be estimated and calculated. Despite the lack of perfection, the results obtained can offer a sense of direction towards eating habits and serving meals. Various sources have listed the number of calories contained in different foods. The essence of using visual analysis is to allow the application can sense and process pictures of meals with a little information about some of the ingredients provided before taking the pictures.

The field of handling visual data experiences challenges like blurry and unclear images. Pictures can contain similar images to those of other items hence leading to wrong feedback from the server. The process of communication from pictures to extracting information from them takes critical stages and various complex attributes must be integrated. The integration process must be keenly programmed to prevent random responses such some of the images can be taken containing many components that cannot come out clear when an analysis is performed (Singh, Dey, Ashour, & Santhi, 2017). The three main objectives of performing a visualization data analysis are: 1) for presentation purposes, 2) for confirmatory analysis reasons, and 3) for exploratory performance analysis. When conducting a visualization for presentation, presented facts are assigned a priori the techniques used for the appropriate presentation chosen depending on the user preference. When choosing a preference of the user, it provides a better ground for effective and efficient communication about the visualization analysis results.

Mobile devices have a heterogeneous life that is very complex in understanding. Huge data is collected with less potential of offering results that are well explained or analyzed with the current tools and apps. People like taking selfies and other photos while having fun and having a good time. In such occasions, different meals are served at different intervals that can lead to over-consumption of calories in the body. Every mobile phone user has different ways of using the device depending on the age and technicality of the life led. While collecting data and valuable information for the analysis and evaluation, the most interesting activities and patterns of life are captured to give the real character of a person. The people who are at the risk of becoming overweight or obese must be cautious about the level of calories intake on their meals (Tomar, 2017). For example, teenagers have a higher chance of becoming obese than the middle age population who are in a marriage and spend much of their time at home. The meals served are balanced and hygienic. The major aim is to determine the number of calories consumed by people than the quality or health benefits derived. Based on the average recommended calories intake per day, the lightweight visual data analysis mobile application must be in a position to extract useful information concerning quantity. While processing the information, short duration events must be taken into consideration since the occurrence is numerous. The use of visualized data analysis by mobile applications have the ability to highlight outliers, shows the trends, indicates clusters, and any gaps existing within the process can be exposed.

2.3.2 Self-Monitoring Visual Data Analytics

Self-Monitoring deal mostly with data collection and feedback, on the other hand, visual data analysis mostly deals with the feedback aspect. The main difference between self-monitoring and visual data analysis is the fact that self-monitoring tools may not always be represented visually, for example in the form of a text. However, visual data analysis offers an interactive visual representation of all the information collected during the self-monitoring process (Chittaro, 2006). The design and development of vibrant yet effective self-monitoring tools is a fast growing industry whereby Human Computer Interface (HCI) experts have integrated such technologies into mobile and handheld computing devices. Now with the push for mobile health applications, the demand for self-monitoring tools is at an all-time high. Such tools are aimed at supporting people’s health and wellness while providing the user with visual feedback to reflect upon his or her behavioral choices. With regard to overall self-monitoring application design. The focus is to reduce the data capture burden on the user by utilizing automated sensors already present in most mobile devices to collect user information (Chittaro, 2006). In a sense, one can say that a lightweight visual data analysis integrated into a mobile device collects statistics that are more accurate and provides the user with real-time feedback. In addition, the visual data analysis and feedback of self-monitoring have contributed in motivating people to capture more information and encourage self-reflection. Major reasons as to why a lightweight visual data analysis tool on mobile devices can get accurate results are due to the use of less memory and small amounts of data to provide results.

Perhaps the most challenging aspect of a self-monitoring tool during the design process is the continued tracking of a user’s statistics through automation (Chittaro, 2006). However, with complete automation, the purpose of self-monitoring is no longer present. Self-monitoring is only effective if users are aware of their behaviors and identifies how to change them. By automating the process, you isolate users from their data. In this regard, visual data analysis can be used to enhance the user’s awareness of their own behaviors and activities as well as providing useful feedback and motivate them to continue with self-monitoring (Choe, Lee, Munson, Pratt, & Klentz, 2013). Overall, more self-monitoring activities and continuous tracking increase the possibility of the creation of an appropriate feedback loop. Thus visual data analysis on mobile self-monitoring devices is a vital element that offers its users with valuable tools to continue tracking their behaviors.

Self-monitoring and visual data analysis tools provide feedback that is supportive of a user’s goals. Studies into the subject further revealed that visual feedback is even more effective when a person’s current state is included in their feedback (Choe, Lee, Munson, Pratt, & Klentz, 2013). Another investigation into the subject showed that when feedback is presented in a certain way it has the ability to alter people’s confidence levels, raising them to enable them to meet their goals. However, you have to take into consideration that varying configurations of feedback will elicit different results from different users. Therefore when planning the design to display self-monitoring feedback, designers need to take into consideration a couple of different designs and how they provoke a different reaction from the various users. This process helps designers to iterate between different design prototypes to select the best fit and one that supports users to meet their weight loss goals and objectives effectively (Kazdin, 1974).

2.3.3 Monitoring Health using Analyzed Visual Data

Obesity has become an increasing problem worldwide with areas such as the United States being the worst hit. According to the Centers for Disease Control (CDC), at least one-third of the United States, the adult populace is obese, that is approximately 35% of the entire inhabitants (U.S. Department of Health & Human Services, 2016). Obesity by itself is not an infection or disease by a gateway to problems such as stroke and heart diseases. A person is considered obese if their weight is more than the maximum weight limit for someone of a particular height. This is known as the body mass index or BMI. The body mass index is a person’s weight (KGs) divided by the square of their height (meters). The higher the BMI the higher the chances of being obese (U.S. Department of Health & Human Services, 2016). As such, a walk to your local chemist or pharmacy you are likely to find a wide array of drugs that offer weight loss solutions, however only a handful of such products provide people with all the necessary knowledge and equipment for a complete lifestyle change (United States Preventative Services Taskforce, 2012).

According to the USPSTF, an effective weight loss program ought to be comprehensive in order to be considered as behavioral intervention technology (United States Preventative Services Taskforce, 2012). Weight loss programs have typically been characterized as behavioral interventions that dictate reduced calorie intake and more activities that facilitate the use of stored energy within the body. Such programs involve setting goals and application of self-monitoring strategies. Through self-monitoring strategies, you can effectively your calorie intake as well as physical activities that help you use up stored energy; being conscious of your own behaviors. Self-monitoring can be defined as an individual strategy aimed at recording, analyzing, and providing useful feedback. The main objective of self-monitoring is to increase or decrease certain aspects of a person’s everyday life (Foster, Makris, & Bailer, 2005). However, in most cases, a person involved in self-monitoring aims to improve his or her individual functionality, academic capacity, behavior. Rather than focusing on the negative aspects, self-monitoring strategies are designed to develop an individual’s skills that would lead to the desired outcome.

According to (Rohrer, Cassidy, Dressel, & Cramer, 2008) obesity is a worldwide problem that should be treated with utmost care and provide patients with the tools and information necessary to monitor their behavioral choices and identify where to make amendments that would lead to effective weight loss. According to Pender, people are more likely to participate if they believe the activity is beneficial to them (Sakaraida, 2010). Therefore, the use of self-monitoring technologies to track one’s body weight, physical activities, and dietary consumption has become a common practice by health corporations and individuals. The increased awareness of these self-monitoring technologies is as a direct result of the need for weight loss programs that are simple and easy to follow through. However, understanding the best way to translate the information that is gathered by some of these self-monitoring technologies proves to be a challenge in the very least. Therefore, as part of behavioral weight loss technological interventions monitoring ones dietary intake is a key aspect of its success. Behavioral intervention technologies or BITs are applications running on common everyday use devices such as tablets, mobile phones sensors, and other mobile devices to support health improvement and wellness activities. From a theoretical standpoint self-monitoring, is the predecessor of self-evaluation, which eventually leads to reinforcement strategies for the changes achieved (Foster, Makris, & Bailer, 2005).

Self-monitoring is essential when trying come up with new behaviors to manage to weight loss goals that include paying specific attention to a specific aspect of a person’s activities and taking note of the main details of his or her behavior. For an effective implementation of self-monitoring strategies, the activities must be recorded with the conditions under which they take place and both their long-term and short-term effects on the subject (Bandura, 1998). A successful self-monitoring endeavor is partially dependent on the subject’s consistency and truthfulness in relation to the targeted behavior, such as calorie intake. Actually, between the years, 1985 and 1990 self-monitoring only referred to following paper diaries with diets written on them. During these years, scientists and dieticians discovered the direct co-relation between weight loss, calorie intake, and physical activity (Jakicic, 2002 Dec). Today, in addition to physical activities and following a diet, self-weighing has also been introduced as another self-monitoring component (Linde, Jeffery, French, Pronk, & Boyle, 2005 Dec).

Although self-monitoring has been highly regarded as the keystone of behavioral weight loss through behavioral intervention technologies there is a whole other side to self-monitoring that is often overlooked. Self-monitoring is only as important as the feedback delivered to its users. Current research has focused its efforts on food databases, intervention methods, and new ways of automatic self-monitoring data collection and analysis techniques. In addition to collecting and analyzing a person’s behavioral data, these technologies need to be able to present this information to the user in a simplified and easy to understand format. However, very little or no efforts have been applied in the research of visualized user feedback and its effectiveness. The norm is that most commercial applications offer simple 2-dimensional feedback with one common piece of information that through self-monitoring activity that is more physical and less calorie intake is best for weight loss. To some extent, this statement holds true, but for vigorous and intensive physical activities; the human body needs enough tie to recuperate and regain all the energy that was used up. When it comes to calorie intake, the situation is even more complicated. The amount of calories alone does not necessarily tell us what entails a balanced diet.

Therefore, there is the need for a more up to date feedback mechanism that allows for lightweight visual analysis of collected self-monitoring information. Information such as nutrients in meals, self-monitored activities, and eating motivators are considerably easier to analyze with visualizations on a mobile device (Fox & Duggan, 2012). It is easier for self-monitoring users to understand their behaviors from the data collected and learn how they can change it for the better, towards be in good health. Mobile devices, presumably smartphones have since become a common device and are widely accessible to many people worldwide. Smartphones have revolutionized the communication industry so mu that their primary purpose is not only for communication. Today smartphones and other mobile devices can be used to store applications for various uses. One such function is related to a person’s health. Such applications can be used for self-monitoring purposes and collect user data, providing them with real-time feedback. Consequently, through advancements in technology self-monitoring applications have developed greatly through empirically testing the overall effectiveness of real-time mobile devices (Fox & Duggan, 2012). Such a tool can be characterized by any number of mobile handheld devices such as smartphones, tablets, body monitors, and so many other devices that have overcome the implementation barrier for mobile self-monitoring devices. The basic operation of these mobile self-monitoring devices is described in three stages:

What people do (behavior patterns)
Why people act the way they do (psychosocial behavior patterns)
When do people act (triggers and behavior timing)

2.4 Data Exploration and Analysis Literature

Information collected from self-monitoring forms an important aspect for self-reflection. Nonetheless, most of the existing self-monitoring applications only have the capacity to show very limited interfaces for the analysis and exploration of data (Kazdin, 1974). With the introduction of mobile visual data analysis platforms, all the aggregated information collected over time provide valuable identity trends as well as averages of their behavioral highs and lows. All the aggregated data provides people with a point for comparing several data trends and identify whether it is positive or negative. In this case, visual data analysis could assist users to identify noticeable trends in their behaviors and activities in an interactive manner.

Most of the self-monitoring tools are either wearable sensors or smartphones. Some of these sensors come with small or no displays at all, which pose a unique problem of having to force the user to install additional hardware to be able to trach their behavior. To address this challenge, some health-conscious organizations have leveraged mobile devices with considerably larger displays such as smartphone interfaces.

Some of the feedback for self-monitoring applications utilizing mobile and hand-held devices is delayed to the user in real time whereas others have to wait until the data collection process is complete. Such discrepancies arise because of the device’s form factor and the nature of data being monitored. A clear example of a device offering real-time feedback include some types of pedometers such as Fitbit, that show a person’s total number of steps sporting a small inconspicuous display. Also, some smartphones also support the installation of pedometer applications that also include real-time feedback. However, other self-monitoring applications such as Jawbone Up Band do not support real-time feedback due to the lack of a display. Such devices incorporate the use of third-party applications to synchronize and view collected data. Real-time feedback comes in handy when one can be in a position to alter their behavior as compared to changing your habit after a long duration of collected data, for instance, hours, days, and weeks can pose a significant challenge to the user.

Among the core objectives of a self-monitoring application is to facilitate the continued tracking of an individual’s behavior. To achieve this objective is a hurdle that must be overcome to enjoy the full benefits of mobile devices fitted with self-monitoring applications. The main problem arises due to the forgetful nature of human beings. Visual reminders provide its users with prominent and effective ways that encourage self-monitoring users to reflect and explore their collected data. These visual reminders can be in the form of notifications on a user’s smartphone, alarms, and emails as well as so many other forms of reminders that are accessible to the user. Also, a widget with a visual summary of all the collected self-monitoring data could take the place of a quick access point and a starting point for data exploration.

2.5 Visual Data Extraction for Analysis

While focusing on the effect of self-monitoring on weight loss, it is measured using a wide array of approaches. Self-monitoring has been considered a favored choice among various clients seeking to engage in weight loss by limiting their daily calorie intake and increasing their daily physical activities. A proven technique has been the self-monitoring intervention through visual interaction with the application.

Regarding users’ data and associated feedback, there are predetermined interventions for particular self-monitoring activities and behaviors. A good example is tracking a person’s daily calorie intake to target a healthy lifestyle with a balanced diet focused on a wholesome nutritional plan is an indispensable aspect of healthy lifestyle choices (Liberati, et al., 2009). These interventions also promote the user’s wellness and reduce the risk of other major illnesses. Therefore, for the overall design of appropriate self-monitoring interventions, you have to take into consideration two major aspects:

The context of the collected information
The overall impression of the data

The context here refers to the values evaluated in relation to the user’s expectations and objectives. Context can be further subdivided into individual and normative content. The individual self-monitoring context is used to describe the user’s values from which collected values can be associated with regard to previously collected information, the baseline (Baker & Kirchenbaum, 1993). On the other hand, normative context can be used to characterize generalized values to which all the collected data values can be compared to help in the interception of collected data values such as a recommendation to complete about a thousand steps each day (Liberati, et al., 2009). The impression in the collected data can be because of two reasons, either impression tracked data, or impressions collected data. With regard to the impression of all the data collected, the input is usually manually entered and to some extent may be incorrect. A good example of tracked data includes the number of calories ingested in a day or the total amount of time you take to complete a certain activity. Conversely, according to (Baker & Kirchenbaum, 1993), contextual data is the imprecise information that users compare values when they are faced with the challenge of an impression of either manual or automatically collected information, such as the activity of coming up with the baseline data. As a result, the information is not well defined and the consequences of deviating from the plan are different and also not clear; for example, it is better to eat one more unit of a vegetable than to add on to the unit of oils and sugars.

2.6 Visual Data Scalability Analysis Review

The form of databases created in the highly advanced technology era has enabled better interactive models of performing visual approaches. Visual analytics is a very sensitive program given priority to handle data mining analytics with complex information. The challenges experienced in the ongoing research of visual analytics can be solved by scalability.

The amount of data determines the level of challenge experienced when handling the data. The user display device and other cognitive limitations concerning the hardware used by a researcher in collecting the data can pose s major challenge. The display elements can have an ability to display less information than the fed data in the system. Closure and continuity are a requirement in conveying specific perceptions of data. Visualized clustering of many elements of containing negative information creates a challenge in the determination or n achieving the intended results. When making use of clusters, doing too much plotting of information can make the results impossible to make judgments because a true distribution of data cannot be achieved. Humans feed on various meals during the day and such kinds of food contain different amounts of calories and other nutrients. When the meals are cooked and served, different images can be seen to come out depending on the process and style of cooking. The major determinant of the outcome image being the recipe followed and the level of expertise. The dimensions used while taking images can be a great contributor to the capability of visualization to be achieved (Kerren, Purchase, Ward, & Dagstuhl Seminar on Information Visualization– Multivariate Network Visualization. 2014). The number of elements in a displayed item contributes greatly towards the scalability of a display. The level of computations leads to the challenges experienced when handling visual information that is used to analyze situations. The example of meals can be having several different images that can lead to different results depending on the manner of cooking. However, the process of fetching the data can be supported with a little selection of the items in the meal taken. Some of the inherent limits to the display capacity of displayed elements are determined by the scatterplots and coordinate matrices. When selecting the data, the user must ensure that every aspect of the selection criteria chosen gas been applied to the required level. During the computation of the visualizations, the complexity of the used algorithms must be core to the computer science topic.

Some of the algorithms can include quadratic efforts. For example, when visualizing many items an interactive experience may be encountered while certain items can take long hours to make updates and process the input data. When dealing with such complex and numerous information, certain approaches can be used like hardware-oriented to aid in the distribution and parallelization of stored data. Some of the items can require software-oriented computational skills to execute commands and filter information. Some of the software-oriented techniques include regression analysis and data sampling. For example, when handling multi-resolution items, the approach of filtering the information must be applied to help separate the parts with high resolutions from the lowly saturated resolutions. Both the computational and visual limitations do not dependent on themselves but they get help from one another. Certain approaches are used to perform the activities of analyzing the data with set specific limits in a simultaneous manner. When large samples are used, the aspect of computation efforts is limited to allow the use of visuals to take effect in handling such complex data.

A display scalability is the ability a data analysis tool has in ensuring the visualization is effective when sent from one person to another using digital assistance to be displayed on wall-sized screens. When we consider the current visualization systems, all designed to be viewed on desktop displays that limit usage on small screens like mobile devices.

A subset of data can be used to offer simple visualizations using the scalability of information. The rate of dynamic change of data and presentation facility scale targeting a specific audience is the general perspective of information scalability. The ability to handle heterogeneous data in several ways is also a form of information scalability.

The human scalability concept deals with the number of human power required to solve an analytical problem. Humans are used mainly to set graceful scaling measures taken from single users to be used in a collaborative setting.

When dealing with algorithms, automation of information can fail to be faster when handling an increasing dataset going parallel with the set computing infrastructure. It is projected in the next 15 years there will be effective and efficient ways of handling visual information that in the present. Computations have challenges such as the use of input of wrong formulae but the extraction and processing of visual objects can be a solution to most of the data.

 

3.1 Visual Data Analysis Tools
3.1.1 Tableau

Tableau offers business intelligence solutions that are preferred to be visualization tools capable of managing visualized interactions within a short period aided by dragging and dropping options. Some of the options data is delivered include heat maps, bubble charts, maps, scatter plots, and pie charts from the information extracted from the dashboard and other diversified datasets. Some of the actions done by tableau include drilling, aggregations, or highlighting in forms of charts helping users to create visualizations that illuminate huge data. When data is sorted in excel, text files, and CSV files can be recognized. A database connector and expertise are needed to extract data from databases. Performance of definitions and calculations can be done by tableau tool. A selection of tableau software offers includes Tableau Server, Tableau Public, Tableau Desktop, or Tableau Mobile to choose from depending on the available resources available with the user (Morton, Balazinska, Grossman, & Mackinlay, 2014). Visualizations can be shared among different users with the Tableau Server because access is restricted to different views that underlies when applying the data by users. The manner a user clicks a mouse within the screen.

3.1.2 Data-Driven Documents (D3.js)

In the recent decades, the use of data-driven documents proved to be among the highest rated in the utilization of visual data analysis where the user data could be sourced free. The best tool for handling a data-driven way of manipulating and visualizing DOM elements must think of using the D3.js tool. The D3.js is a JavaScript element known for data-driven documents makes use of in the manipulation of documents where the provision of data is provided by the researcher (Nair, Shetty, & Shetty, 2016). The raw or unprocessed data is analyzed and the output generated with the help of SVG, HTML, and CSS. Some of the images produced by the D3.js tool are as shown below.

 

Figure 3. 1 Bubble Chart

Figure 3. 2 Scatterplot Chart

 

Table 4. 1 D3.JS vs. TABLEAU

D3.js
Tableau

The JavaScript Library found in the D3.js does not have a visualized software
It is a business intelligence software that includes a visualized software package

It is open source and free
Very expensive and proprietary

The learning is very difficult because of the heavy coding required
An easy to learn a tool that uses a drag and drop option

The time required to use tableau ranges from hours to days
The time required to use D3.js is minutes of navigating and dropping items on the dashboard

The output from a D3.js is done in a scalable vector graphics (SVG)
The visualization format used in tableau can allow exports in EMF, JPEG, BMF, and PNG formats.

When the data used in gigabytes, there is so much struggle experienced during the processing period.
A tableau is an able tool that can identify measures and dimensions. Tableau also has the ability to handle gigabytes of data.

3.1.3 WebDataRocks

Getting an efficient tool for performing a data analysis with a combination of visualization becomes a challenge for many researchers. Using WebDataRocks helps in enhancing accuracy, offering efficient display, and making data extracted from JSON and CSV files. WebDataRocks is a form of pivot tables operated by the web that allows users to visualize given information in a better way. The aggregate and insight received from a WebDataRocks tool occur in real time. The major use of a WebDataRocks is to sort, give an average, and count the given records issued from the primary source to offering a summarized data in form of a grid. The best web reported is experienced by the delivery of reports from the JavaScript tool. WebDataRocks offers the best reports that can be used universally with many users from different industries (Huerta-Cepas, Serra, & Bork, 2016). The data used to generate the reports must be provided by the user depending on the topic of study or the location of the population sample. The most achieved advantage of making use of the WebDataRoks is the ability to provide numerous data analysis and examination features making the web reports to users easier to interpret and understand the generated results. The WebDataRocks does not need to be deeply understood with the latest high technology in the industry. Once the users are able to load and run JSON or CSV files the start of reports generation can take place. The dicing and slicing of data are done by dragging and dropping fields, drilling them down, filtering, and sorting. The WebDataRocks tool can be integrated with the Angular framework.

3.1.4 BIRT

BIRT is a tool used for the creation and generation of visualized reports that have the capability of being embedded in other web-based applications. BIRT is a free tool found on the open source learning materials that researchers can make use while performing experiments on data visualization techniques. BIR offers the support of Java EE and Java applications in special ways that can help to generate quick and detailed reports from the input data. Unlike other visual data analysis tools, BIRT is built with two major components that create a difference. The components are a runtime element that is able to generate the designs that can easily be deployed in any Java-enabled environment and a visual report designer that helps to create different designs according to the tool requirements and ability (Huerta-Cepas, Serra, & Bork, 2016). The two listed components are the core principles that help BIRT work well. However, the tool is able to allow a charting integration engine to be installed into BIRT designer. The data input to a BIRT visual data analysis tool must be generated from specific sources that are compatible with the BIRT data analysis tool. The acceptable sources for BIRT data information include POJOs, Web Services, JDO data stores, XML, JFire Scripting Objects, and SQL databases.

3.1.5 Google Charts

Google has been a very powerful tool in data analytics, visualization, and reporting. The same applies when working with Google Charts. The tool is far much better beyond being effective. The tool is free among the open source items and designed in a simple way of working that cannot give users a difficult time to use. The form of gallery generated by Google Charts is rich in customization of results based on the user preferences. The controls give multiple controls that can display dynamic data and support the portability and compatibility of cross-browsers (Huerta-Cepas, Serra, & Bork, 2016). With the latest technology trends, Google Charts are popular in the market for visualizing data using online tools.

3.1.6 Cytoscape.js

Cytoscape.js is another free or open source tool for data analysis that can perform visual data analysis. The language used in the tool is JavaScript. The tool is able to provide a library of visualization and graph analysis theories. Cytoscape.js is highly rated among the best tools for with high efficiency in the market in processing and manipulation of data into interactively displayed graphs (Franz, Lopes, Huck, Dong, Sumer, & Bader, 2015). The Cytoscape.js can also be integrated into an app of choice by the user.

3.2 Mobile Device Visualization Approaches

Visualizing information is a technique used when the alternative automation tools for data analysis become complicated and generated results fail to work as expected. The use of numeric data creates decision-making opinions that can be interpreted from the results that can help in determining various aspects. The use of data visualization highlights crucial information while hiding non-required credentials. When the users of the analyzed data have a direct exploration with the information, the findings are achieved faster with high confidence levels. The mobile technology is an available tool with everyone currently in the world that can capture data in various forms. Taking pictures and recording data being the primary aspects. Using the data collected, created applications can be used to perform a data analysis. The activities engaged with mobile devices require less engagement time and the generation of results must be produced almost immediately. Tasks requiring long periods of performing the analysis can be well done using a desktop but mobile devices can be well used to monitor instant activities such as the collection of information from served meals or the number of steps made by a person within a day. The extension of users’ decisions when and where they are needed has been facilitated by the presence of mobile devices. A contemporary system has been made possible by the mobile components. While handling daily tasks and fulfilling assigned duties, mobile phones can be used to run various applications and generate results without affecting the time needed in other activities. The same case such as when a person received an urgent call and responds to it, the examination or analysis of visual data can be done within the minimal time that cannot hinder certain activities from taking place. The technical limitations of mobile devices applications have rapidly aided the development of mobile activities and tasks performance in a significant way. The specific features that have made mobile devices have an upper hand in the examination of data are the screen resolutions that are extended to 1920 x 1080 pixels.

Very intuitive and technical interaction methods are required to supplement the soft touch screen keys. The small spaces provided by the mobile devices must be maximized with visualizations with compact results. The mobility of the devices challenges the designing of mobile interactions. The loudness and other lightening usage context make computers more stable than mobile devices. The advantage of physical rooms comes when recurring activities but mobile devices are the best when working in highly variable conditions. The variable conditions are characterized by implications with graphics that can be perceived in dark and light conditions. The different changes in the environmental conditions must be taken into consideration. The convenience, ease of use, and availability make mobile devices a better gadget in performing data analysis. The same manner some of the certain activities having a routine, mobile devices can be included in the routine activities such as being constantly connected and offering supportive information effectively. There exist embedded sensors on mobile devices that are not present in desktops. Some of the sensors include pedometers, physiological sensors, geographical positioning, light, accelerators, and proximity.

3.2.1 Approaches of Compact Visualizations

The act of visualization works on how the space on the screen is utilized. To perform better exploration of huge and large documents in an effective way, desktops are used to offer different perspectives of views. The summary provided gives a faster facilitation of the required access to content. Various navigation and presentation techniques have been devoted to working with 2D data and extra-large documents used by desktop systems. The difference in the visualization can be detected from the summary produced. To make the visualization process simple, sort of traditional solutions have been set to counter the problem and they include:

Detail and overview approaches
Information page restructuring
Zooming and panning/scrolling techniques
Context & focus approaches
Off-screen objects or contextual visualizations

However, information space restructuring applies universally on both mobile and desktop applications. A manual form of designing web pages to fit every device targeted enables better use of data approach. Apart from manual designing, reformatting automatically offers a possible solution. The original layout can be preserved by the compression of the available space into a thumbnail.

3.3 Cutting-edge Visualizations in lieu of Mobile Interaction

In this section, we get a better understanding of mobile device abilities that can be made use to enhance and improve the current performance of mobile devices. The manner of performing tasks can be facilitated and sped up by highly interactive and easy to use systems are implemented. Due to the limited space on mobile devices, solid methods of data extraction and processing are needed. The use of mobile devices can be set to collect a user’s eating behavior and perform an analysis. A behavior ring is produced with solid visualization where timing can be tracked.

Despite the known advantages of self-monitoring devices, most of these devices have a high rate of non-wear, which ultimately leads to misleading feedback due to missing data. Also, these self-monitoring devices are usually unable to accurately capture user behavior. However, it has been noted that most users of self-monitoring devices have at least one smartphone or access to one with the capacity to operate as a self-monitoring tool (Finkelstein, Trogdon, Cohen, & Dietz, 2009). This paper aims to investigate the design, development, implementation, and use of mobile self-monitoring applications. In addition, how it has affected aspects such as the collection and analysis of data as well as how feedback is relayed to the user (Finkelstein, Trogdon, Cohen, & Dietz, 2009).

The paper will use the example of a smartphone application or ‘app’ as is commonly referred to, that combines self-monitoring strategies using sensors inbuilt within the phone and a user’s goals and objectives. Therefore, the self-monitoring application either makes use of a smartphone’s built-in sensors to manually or automatically collect a user’s activity and behaviors. Also in addition to the likelihood of non-wear being drastically reduced to levels that are more manageable. After the data collection is complete, the smartphone has the capacity to trigger an analysis and evaluation of the data and provide feedback in real time (Webber, Tate, & Quintiliani, 2008). At the end of the day, the self-monitoring application will allow users to interact and review all the collected data and effectively alter their behavior and physical activity, such as reducing calorie intake and by how much.

3.4 Dietary Intake and Its Contribution to Weight Loss through Self-monitoring

Obesity is a major contributor to the continued rise in spending on healthcare products. According to the Annual medical spending attributable to obesity: Payer-and Service-specific estimates between the years of 2006 and 2008, the annual spending on weight loss products shot from forty million dollars to well over a hundred and forty million dollars. It has been predicted that should this trend continue unhindered by the year 2023 more than 80% of U.S citizens will be overweight (US Department of Agriculture and U.S. Department of Health and Human Services., 2010. Dec). As a result, of most self-monitored weight loss programs are aimed at altering the activities and behaviors of users for a better lifestyle. Some of these interventions include:

A protracted and continuous intervention contact,
Self-monitoring strategies
A sense of accountability
Motivational cross-examination
Regular self-assessment
Consistent physical activity

The American Dietetic Association (ADA) has concluded that by realizing a negative energy balance is one of the most significant factors that directly affect the extent and overall rate of weight loss over a given period of time (US Department of Agriculture and U.S. Department of Health and Human Services., 2010. Dec). Some of the strategies that are used to ensure the achievement of a negative energy balance include:

Calorie counting,
Modifying macronutrient composition
Manipulating a meal’s energy density
Opting for low-calorie diets

Also through the reduction of dietetic fat and starches is a more applicable technique that can be used to reduce the total calorie intake by between 500 to 1000 kilocalories (kcal) each day, which directly translates into one to two pounds per week aiming to lose weight. It has also been noted that however small a drop in calorie intake when it is combined with an increase in a user’s physical activity, will result in a clearly visible weight loss. This technique has a higher chance of implementation and an even greater possibility of being sustainable in the end.

The U.S. Department of Agriculture has also taken the initiative to help people better understand and interpret dietary guides and self-monitoring strategies, and in essence, it promotes a diet discouraging high fat intake and encourages participants to eat fruits and vegetables. The general communication you can pick from all this is that an appropriate nutritional habit has the ability to promote a person’s health and diminish the risk of getting a chronic disease. According to the guidelines put across by the Dietary Approaches to Stop Hypertension (DASH), the typical diet should be a representation of healthy living by promoting the ingestion of vegetables and fruits, low-fat milk and milk products, whole grains, reduced consumption of meat. By the end of it all, you will have cut your total calorie intake by approximately 30% (United States Preventative Services Taskforce, 2012). The proposed diet plan in conjunction with self-monitoring has proved to be effective. More so, the pursuit of weight loss has been managed by self-monitoring tools own and managed by an individual that help in tracking various activities.

3.5 The range of Self-monitoring mobile device application

Many self-monitoring applications are developed to cater to the target market of both health care providers and individuals at home. These types of applications vary in complexity and application but in general, they are more adapted for everyday use by untrained individuals. However, some of these applications may exhibit some forms of common medical terminologies and utilities that may or not be understood by non-health professionals. In a study published in 2012 (Conn, 2012 Dec), a group of individuals both with and without healthcare backing indicated that the most common category of mobile applications are clinical decision-support tools and medical education resources both of which border closely to self-monitoring especially when applied to weight loss and self-diagnostic scenarios (Bandura, 1998). This category of applications can be regarded as patient-centered applications each capable of accomplishing a wide array of functions such as manages chronic illness, lifestyle intervention, and self-diagnosis. These types of applications incorporate a variety of functions and usually log user data such as daily eating habits, total calorie intake, compliance with medical procedures, individual physical activity and their behavior in a mobile offline/online database.

A large number of applications in this category have focused on exercises and weight loss. A smartphones inbuilt camera, which has become somewhat of an industry standard has the capability to capture images and record a photographic diary of daily calorie intake with regard to food and drinks ingested.  Most self-monitoring applications utilizing the portability and ease of access of smartphones today are used to track calorie intake and weight loss goals by objectively tracking daily physical activity from sensors within the phone such as pedometers and accelerometers (Conn, 2012 Dec).

3.6 An Investigation of the Effect of Mobile Devices on Self-monitoring

Among the most noteworthy recurring hurdles and challenges in the self-monitoring application field is the continued need for effective and consistent processes for measuring user behavior and physical activity for the purpose of scrutiny and intervention for continued health benefits (Finkelstein, Trogdon, Cohen, & Dietz, 2009). There is growing apprehension over the justification of mobile self-monitoring applications as a direct result of the possibility of errors, misinterpretation, and bias. Especially among the youth who have adopted the use of “objective” measures of physical activity and behavior, such as heart rate monitors, accelerometers, and global positioning systems (GPS), all of which are available on their smartphones, are more likely to misinterpret the information gathered. For example, more than a few researchers have found inconsistencies in their levels of physical activity and associated behavioral patterns when comparing self-monitoring feedback using objective assessment methods (US Department of Agriculture and U.S. Department of Health and Human Services., 2010. Dec). However, presently most of these objective self-monitoring activities are being deployed in large-scale as a means of investigating an adolescent’s behavior and physical activity with the promise of obtaining a more truthful valuation of one’s physical activity and behavior.

3.7 Self-monitoring and Visual Data Analysis of Dietary Intake

Almost all the studies that were aimed at investigating nutritional self-monitoring have identified a substantial relation between self-monitoring strategies and weight loss goals. Some studies use paper diaries and others used some deviation of the paper diary and a supplementary smartphone application (Jakicic, 2002 Dec). The quantification and subsequent analysis of nutritional self-monitoring vary. For instance, with regard to some studies, the users were trained to take note of their physical activity, disposition, eating environment, water intake, and other important behavioral variables that directly affect the total calorie intake daily. The amount of self-monitoring applied in these participants comprised of recording and analyzing only 5 variables. Therefore, when assessing the result of self-monitoring strategies, the researchers used these dietary variables to determine a monitoring index (Jakicic, 2002 Dec). However, in these self-monitoring endeavors, those with complete records are able to see a considerably larger difference in terms of losing weight. Moreover, the results were even higher for users with higher self-monitoring fulfillment described by (Yon, Johnson, Harvey-Berino, Gold, & Howard, 2007) who assessed the benefits of self-monitoring is directly related to the frequency of self-monitoring.

The introduction of smartphone technology, with access to the Internet, for use in self-monitoring, has come up with a new generation of strategies. It has since been reported that the number of self-monitoring programs delivered over the Internet was more so associated with weight loss. Yon related the results of a weight loss study that incorporates smartphone technology in self-monitoring to a prior study that used traditional self-monitoring and found significant variances in the total weight lost in direct relation to the user’s adherence (Yon, Johnson, Harvey-Berino, Gold, & Howard, 2007).

3.8 Self-monitoring and Visual Data Analysis of Physical Activity

Regarding the utilization of traditional diaries to record one’s physical activity and behavior only, a few of these endeavors scrutinized the part played by self-monitoring strategies in when the question of weight loss is raised. The users who participated in the self-motivation and assessment were asked to take note of their daily routine in relation to the type of exercise and their duration. Self-monitoring was accurately defined by the total time taken attending to certain physical activities were completed. The results clearly depicted that those participants adhered to the consistency of self-monitors of their behaviors and physical activities not only achieved significantly better results but also faced fewer challenges along the way.

3.9 Self-monitoring and Visual Data Analysis of Weight

Only recently have researchers backed weight loss self-monitoring as a plausible solution that could possibly increase a users’ cognizance of their weight, its relation to the number of calories ingested, and the physical activities required achieving the desired weight loss goals. One researcher directed expressive subsidiary studies to two continuing tests by using a solitary item survey to evaluate the occurrence of self-weighing among the participants of the trial (US Department of Agriculture and U.S. Department of Health and Human Services., 2010. Dec). As a control measure, a trial of weight gain deterrence and a weight loss trial were administered at three points. In the weight gain experiment, users who weighed themselves daily often claimed to have lost weight, and the less frequent incidents of self-weighing were linked instances of weight gain. Nonetheless, concerning daily, weekly, and monthly self-monitoring, instances of weighing were related to loss of weight and even more, recurrent weighing was symbolic of an even greater achievement with a 24-month self-monitoring weight loss (Foster, Makris, & Bailer, 2005).

Two completely random experiments aimed at addressing instances of daily self-weighing within a self-monitoring and regulatory framework as the main strategy to determine the effect of weighing among these three groups of people: they include face-to-face, internet-based, and an 18-month trial fixated on the deterrence of regaining weight. The results showed that both groups gradually increased their day-to-day weighing, which was expressively connected with a lower possibility of regaining weight. Therefore, the user’s observance to weighing progressively lessens over time in both groups (Choe, Lee, Munson, Pratt, & Klentz, 2013). When comparing the results of user behavior and self-regulation strategies within a timed clinical experiment. Members of one group were asked to record their daily weight using a digital weighing scale. ON the other hand members in the second group got a modified behavioral treatment strategy whereby they were directed not to record their weight until the 11th week and then to subsequently weigh themselves weekly.

After 20 weeks, the frequency of weighing was expressively associated with weight loss; however, there was no noteworthy variance in the weight loss margins for both groups. The use of an electronic scale offered its users feedback in the form of objective data to ratify the self-reported devotion to weighing, which was usually over 95% of the days the experiment was live.

4.1 Quantitative Visual Data Analysis Using Regression Models

Computation and interactive data visualizing analysis methods offer very useful integration and study of data in a successful manner. When doing an interactive data analysis workflows and explorations, the quantitative means are used in externalizing the outcome of input data from the research. The use of quantitative data analysis was considered and the regression models applied in the generation and interpretation of the results. The given dataset by users is used to generate results from the models selected. A numeric coefficient benefit is also provided to the user for a better understanding of the outcome. Depending on the availability of the models chosen, subsequent workflow steps can be used in the fulfillment of the tasks. A reconstruction of complex models is then performed using the inversion of applied local models to break down the complexity involved. A sample population of 30 respondents was examined and an average calorie count from the meals served was recorded. The performance of a statistical analysis was taken and the results indicated that a majority of the people had a lot of calories intake during the lunchtime meal. The results from the sampled population revealed that 40% of the calories consumed in a day came from the meal taken at lunchtime. Using the quantitative descriptive statistical results, the mean calorie consumption per meal indicated: breakfast 286.5 calories, lunch 530.5 calories, dinner 430.3667 calories, and snacks 88.1333 calories. The data collected from the respondents and statistical results are as indicated below. The recommended calorie intake for men should be 2500 while for women 200 calories. However, the research indicated instances, where every responded, was consuming fewer calories than the recommended amount. The reason why such little calorie intake could sustain a healthy living could be a possibility of lack of exercise. The calories consumed are not burnt hence maintaining the calorie count balance.

 

Table 4. 2 Sample Population dataset

Calories Counter

Meals

Respondents
Breakfast
Lunch
Dinner
Snacks
Total

1
300
528
412
92
1333

2
297
538
428
87
1352

3
312
546
450
81
1392

4
289
512
467
93
1365

5
309
516
432
91
1353

6
311
572
419
83
1391

7
299
564
463
89
1422

8
293
513
443
81
1338

9
271
502
427
82
1291

10
315
512
429
92
1358

11
272
519
434
90
1326

12
269
540
439
87
1347

13
298
532
417
82
1342

14
291
529
438
93
1365

15
289
516
428
94
1342

16
268
503
421
86
1294

17
301
537
431
87
1373

18
250
512
409
88
1277

19
267
517
416
92
1311

20
238
523
422
94
1297

21
289
561
420
97
1388

22
273
551
431
89
1366

23
248
539
426
85
1321

24
315
542
428
94
1403

25
264
526
423
84
1322

26
319
537
432
86
1400

27
283
520
430
82
1342

28
271
526
431
83
1339

29
294
548
429
91
1391

30
300
534
436
89
1389

Total
8595
15915
12911
2644

Average
1351

Table 4. 3 Descriptive Statistics

Breakfast

Lunch

Dinner

Snacks

Mean
286.5
Mean
530.5
Mean
430.3667
Mean
88.1333

Standard Error
3.8856
Standard Error
3.2241
Standard Error
2.3525
Standard Error
0.8329

Median
290
Median
528.5
Median
429
Median
88.5

Mode
289
Mode
512
Mode
428
Mode
92

Standard Deviation
21.2826
Standard Deviation
17.6591
Standard Deviation
12.8854
Standard Deviation
4.5617

Sample Variance
452.9483
Sample Variance
311.8448
Sample Variance
166.0333
Sample Variance
20.8092

Kurtosis
-0.4521
Kurtosis
-0.1578
Kurtosis
2.1413
Kurtosis
-1.0917

Skewness
-0.4850
Skewness
0.5299
Skewness
1.1909
Skewness
-0.0200

Range
81
Range
70
Range
58
Range
16

Minimum
238
Minimum
502
Minimum
409
Minimum
81

Maximum
319
Maximum
572
Maximum
467
Maximum
97

Sum
8595
Sum
15915
Sum
12911
Sum
2644

Count
30
Count
30
Count
30
Count
30

Figure 4. 1 Graphical Representation of Results

Figure 4. 2 Calories Counter

 

4.2 Surveys on Human Effect Caused by Handheld Mobile Devices

The handheld mobile devices are modified to make the use of the latest computing capabilities like video, e-commerce, internet communication, and information retrieval. The incorporation of such features has made the mobile devices so popular among the people with many expected tasks to be completed using the devices. Every social site has developed apps compatible with the devices to allow everyone who owns them to get access to the world platforms. With the many activities and tasks that can be completed with mobile devices, the time spent on the gadgets has an impact on the health of users. Upon a survey done during the research of this project, a survey was conducted that was taken part by 5 females and 5 males. The participation was conducted by answering a set of four prepared questions. The region surveyed was in the New York City. A confidentiality preservation was agreed to take place and the list of participant names was issued with alphabetical letters to enhance the privacy and confidentiality of personal information. The summary of responses given was consolidated as below:

Table 4. 4 Summarized Demographic Data

Respondent
Age
Hours spent on a mobile device

A
56
1

B
39
3

D
36
3

E
15
4

F
27
4

I
42
4

C
39
5

J
35
5

G
28
7

H
18
12

5.1.2.1 Questions

How long and how often do you make use of your mobile device?

The spread period the respondents spend on their mobile devices ranges between one hour and twelve hours a week. The collected responses were presented in the chart below.

Figure 4. 3 Quantity of hours spent on hand held devices (HHDs)

 

What is the purpose of your mobile device: entertainment, conversation, texting?

The listed three purposes are used by the respondents but during the interview, we only required the feedback to be given based on prioritized activities. A great number of participants use their mobile devices for texting, others use for making conversations, and the least number of respondents use the mobile devices for entertainment.

Table 4. 5 Purpose of mobile device

Respondent
Age
Purpose

A
56
For conversation

I
42
For conversation

J
35
For conversation

E
15
For entertainment

H
18
For entertainment

B
39
For texting

D
36
For texting

F
27
For texting

C
39
For texting

G
28
For texting

What discomforts have you encountered on your hand or shoulder while using mobile devices?

Among the ten respondents, nine of them confirmed to have experienced hand and shoulder discomfort after a long period of making use of the mobile devices.

Table 4. 6 Discomforts encountered

Respondent
Age
Hours
Purpose
Hand and shoulder discomfort

A
56
1
For conversation
No

B
39
3
For texting
Yes

D
36
3
For texting
Yes

E
15
4
For entertainment
Yes

F
27
4
For texting
Yes

I
42
4
For conversation
Yes

C
39
5
For texting
Yes

J
35
5
For conversation
Yes

G
28
7
For texting
Yes

H
18
12
For entertainment
Yes

Do you ever experience any tangling sensation in your hand or shoulder when you are done typing on a mobile phone device?

Seven out of the ten respondents acknowledged having experienced a tangling sensation in their hand or shoulder after typing for long.

Table 4. 7 Tangling sensation

Respondent
Age
Hours
Purpose
Hand and shoulder discomfort
Tangling sensation experience

A
56
1
For conversation
No
No

B
39
3
For texting
Yes
Yes

D
36
3
For texting
Yes
Yes

E
15
4
For entertainment
Yes
No

F
27
4
For texting
Yes
Yes

I
42
4
For conversation
Yes
No

C
39
5
For texting
Yes
Yes

J
35
5
For conversation
Yes
Yes

G
28
7
For texting
Yes
Yes

H
18
12
For entertainment
Yes
Yes

4.3 Discussion

The users of the developed applications have no idea on the methodologies and programs used in creating the systems but the major objective is to simplify the processing of information. The ability of mobile devices to read or scan codes on products in malls can be applied to generate information from images. However, the images have no hidden codes thus creating a difficult situation in extracting information from pictures. Despite the high sensitivity and data extraction processes needed, the advancement of used resources must be thought about to help the project become a success. The mobile devices can be utilized to offer detailed information but the effects are severe. The analysis done above indicates that tangling experiences and discomfort are a major problem experienced by users. The positions that are taken while using the devices for long can lead to unwanted feelings such as neck pain. The investigations and deductions portrayed in this paper are because of the application of scientific methods with regard to self-monitoring. Self-monitoring is a strategy that is applied to enable the user to be more aware of target behavioral goals and physical activities. Each investigation into the field of self-monitoring is seen to be consistent and usually allied with weight loss programs (Sakaraida, 2010). For the reason that of the inconsistency of self-monitoring diets and physical activities were quantified; it was however not conceivable to account for the exact occurrence of self-monitoring strategies that translated to a discernable difference in the overall weight outcomes. Again, owing to the weighing investigation, there are substantial weight loss variations between users who weigh themselves on a daily basis, those who weigh themselves on a weekly basis, and those of whom weigh less frequently. This outcome was confirmed through the systematic review of the self-monitoring literature.

Most of these participants who were included in the studies through the implementation of a descriptive design to display the procedural flaws through clinical trials of self-monitoring strategies. These restrictions have greatly influenced the level of indication and thus affected the deductions and successive endorsements that can be received from this study of self-monitoring strategies on mobile devices. The sturdiest opinion was the reliable sustenance of self-monitoring strategies on various smartphone platforms in the experiments that covered the self-monitoring review period. Conversely, for the reason, that of the consistency of the participating samples, the generalization of collected information and visual feedback based on findings and data analysis was restricted to overweight or obese people (US Department of Agriculture and U.S. Department of Health and Human Services., 2010. Dec). This signifies among the main limitations in the consideration of the capability, observance, and influence of self-monitoring strategies, visual data analysis, and presentation of real-time feedback over the vast smartphone network worldwide. This also dictates on areas that future research endeavors into self-monitoring strategies should to focus.

An added organizational flaw of the assessment of strategies was the valuation of self-monitoring and allowing for enough room for a quantification bias. With the exemption of the initial studies, that used analysts to evaluate traditional paper diaries on the various self-monitoring activities (such as foods eaten, time these foods were eaten, the quantity of food and even the total calorie intake) and a current experiment that is aimed at defining self-monitoring observance. Though none of the investigations reported such criteria, whereby they assessed self-monitoring strategies and how they identified the completeness of their paper diaries or logs. The investigations showed that participants described taking note of calorie intake on certain days when the paper diary was never even opened and no documentation was taken due to the misconceptions of a self-reported paper diary log (Bandura, 1998). In addition, the use of technology and mobile electronic devices has greatly improved the self-monitoring conduct provided an objective authentication of any self-reported behaviors of interest.

Even though there were procedural restrictions of the investigation of self-monitoring strategies used worldwide, there was sufficient supporting evidence and proof for the reliable and noteworthy constructive interdependence that exists between a self-monitoring the dietary plan, behaviors, physical activity, weight loss programs and their subsequent successful outcomes associated to weight management practices. The investigation carried out on self-monitoring strategies acknowledged several loopholes, including the ideal occurrence and period a self-monitoring diet and physical activity plan, is effective.

The survey was done on mobile device usage and meals consumption have revealed that people make use of mobile devices all the time and meals are taken every day at different intervals. The two activities recur in the lives of people hence finding a way to analyze them and without imposing restrictions can make it easier to understand, the challenges faced in the world regarding health issues. Most of the chronic diseases have been managed by the development of simple mobile gadgets that can be operated by patients or an individual in performing self-tests. An example of such an analysis tool is the kit made for diabetic patients. The sugar levels can be determined by patients and the recommended medication taken or get a quick communication done to healthcare providers.

The examined situations were to provide evidence on the manner information could be analyzed using various tools. Apart from the use of the quantitative technique, a qualitative analytical method of analyzing visual data using mobile devices can be applied. When implementing the qualitative approach, the following aspects must be taken into consideration:

Classification field codes and codes, interview scripts, or observations by the use of inferring images that are intended to determine what is significant.
Performing an examination of aforesaid categories of information to identify existing relationships between the analyzed items
Making of explicit differences, patterns, and commodities
A formalization of theoretical constructs that help to make inferences from related cases in time and place.

 

5.1 Conclusion

The research performed on the ability of mobile devices such as Android phones having the ability to perform various tasks reveals that more can be done with the simple devices. When most of the devices users do not understand the much such phones can do, it is high time for the manufacturers to include useful applications such as self-monitoring tools that can advise on nutrition, and other health-related aspects from the food consumed. In the present time, people with internet-enabled mobile phone devices have mobile wallet payment accounts that allow them to make payments for the items purchased. In the process of making payment for the items bought, the information about the purchased food can be input into the devices and pictures of the same taken for analysis before eating. Upon the various investigations performed on the ability of mobile devices to collect and analyze data, the research on the analyzation of meals using pictures has not been completed and more information and tests are put into action to ensure that the ability of pictures to extract the required and exact details about ingredients can be possible. The major challenge facing the project of visualized data analysis on mobile devices is the outcome of different meals in different forms depending on the recipes followed. Apart from the food recipes, the cooking methods can contribute to different outcomes of the same food cooked by different people. When the result is different, the pictures taken for analysis will different hence giving wrong results. As a prototype form of analysis, the generation of results is currently aided by the use of manually keyed in information that can help the programmed application to compute the number of calories in a specific meal. When the used ingredients are submitted into the application and finalized by taking a picture, eh information contributes much to the determination of the required outcome. The input of such information has been simplified by creating a list of ingredients that can be used in cooking various meals, so the person performing the calorie count calculation will take little time to select the items from a list then quickly take pictures of the meal.

A major challenge facing the effective use of mobile devices is the size of the screen and the ever-changing user input interface. The amount of information expected to be displayed on the mobile device screen is much more than the screen can accommodate. For example, the display of graphs and other statistical representations can be ineffective due to the limited space on the screen where some information might be invisible. The consideration of the screen size opens the discussion about the challenge most people are facing with eye complications due to extended hours f working on computers and laptops screens. Due to the increase of eye problems, many people cannot be comfortable to make use of the small-sized mobile devices in analyzing food data taken from multiples of images and selected data input. The screen is a major hindrance from the success of viewing results on the screens of mobile devices, a major analyzing tool and knowledge must be innovated to counter the challenge.

5.2 Future Research

While working on means of bettering the system or program used to generate results on mobile devices, simple visualization tools must be used than reports or responses. The current stage of giving feedback on the input information does not satisfy the trend in which the analyzed data flows. Representations such as graphs and charts can perform better because they do not require long sentence explanations to offer a recommendation, but the trend of such graphs and charts can be quick and easy to understand. For example, when bar graphs are used, the shortest indicates a less consumption of a product but a long bar indicates too much of something. The future works can concentrate on making the captured data come out as charts, graphs and other forms of visuals that can be seen and interpreted without having to be explained. Visuals are easy and quick in conveying concepts in a non-biased manner because even the illiterate can be able to understand what diagrams or graph mean when displayed on the screen. Apart from offering the details about the subject matter, data visualization has the ability to perform many things when well implemented with proper standards of displaying captured information. Some of the things data visualization can do when the required improvements and upgrades are made include;

The prediction of revenue for businesses, institutions, and organizations following the graphical representations from the past periods
The clarification of factors affecting certain behaviors including customers, politicians, doctors, nutritionists, and many other fields provided the required data is provided and well executed by the set programs on the installed applications.

The visuals to display must be well decided according to age and preference. Various age groups need to view the results presented by the visualization analysis tools in different ways and presentations. To enable consistent use of the applications among various groups of people, the determination of outcome must be done based on the age limit. For example, children must be included charts and graphs that can show on the screen and sing songs of praise to anyone who eats healthy.

Visual data analysis requires groundwork that can offer guidelines on the manner of presenting feedback. Since the project is a new technology in the market that will enable the capture of pictures and generation of feedback by extracting information from images, there must be clear steps towards such an achievement to help the project become a success. Solid data is a basic requirement in the processing and analyzing of information in offering the needs of the audience. The various aspects of consideration when determining the organization of data visualization technology include:

Have an understanding of what is to be visualized. The little information required to offer guidance in the generation of the intended feedback must be supplied into the system.
Understand and know the form of information processing by various audiences. The elderly people do not bother the color outcome but the message contained in the response given by such applications.

The visuals used must be easy and simple to understand the information issued. Creating complex results can demoralize users from taking part in the use of such tools for data visualization because of the challenges experienced in trying to understand the results.

The data analyzed must be understood by the application users to avoid using the wrong data sources. For example, the cardinality and size of the collected data must be limited to generate values in at most two columns due to the small size of mobile devices screen.

Instead of using mobile devices to take pictures of objects and items without getting results from the contents of images, the research will aid in getting a better understanding of the environment and products used on a daily basis. Fitness applications have been used to encourage people to exercise and count track the milestones taken each day. The same is assumed to take place with the visualization of events and activities. Despite the entire idea revolving around in the food sector, future improvements and advancements can be made to include tracking of various activities using images. Technology is a master in the simplification of things and the creation of gateways towards achieving a better world through innovation. Traditionally, pictures were considered to be drawn by artists on paper, but with time, digital photos can be taken and stored for a longer period. The idea of using visuals in capturing and processing data has a great impact on the lifestyle of the people because most of the things are better explained through mages than information or data. For example, explaining the contents of three plates of food all with different kinds can be a challenge but a picture can explain well and offer firsthand information because the information cannot be altered as it happens with recorded data figures. Therefore, a lightweight visual data analysis performed on mobile devices can create a big difference in the lives of people.

References

Baker, R. C., & Kirchenbaum, D. S. (1993). Self-monitoring may be necessary for successful weight control. Behavioral Theory, 24:377–394.

Bandura, A. (1998). Health Promotion from the Perspective of Social Cognitive. Psychological Health, 13, 623–649.

Chittaro, L. (2006). Visualizing information on mobile devices. IEEE Computer, 39, 40–45.

Choe, E. K., Lee, B., Munson, S., Pratt, W., & Klentz, J. A. (2013). Persuasive performance feedback. The effect of framing on self-efficacy. AMIA Annual Symposium Proceedings, (pp. 825–833).

Conn, J. (2012 Dec). Most-healthful applications. Modern Healthcare, 10;42(50): 30-2.

Finkelstein, E. A., Trogdon, J., Cohen, J., & Dietz, W. (2009). Annual medical spending attributable to obesity: Payer-and service-specific estimates. Health Aff (Millwood), 28: w822 – w831.

Foster, G. D., Makris, A. P., & Bailer, B. A. (2005). Behavioral treatment of obesity. American Journal of Clinical Nutrition, 82, 230S–235S.

Fox, S., & Duggan, M. (2012). Mobile health has found its market: Smartphone owners. Retrieved from Pew Research Center, Pew Internet and American Life Project: http://pewinternet.org/Reports/2012/Mobile-Health/Key-Findings.aspx

Franz, M., Lopes, C. T., Huck, G., Dong, Y., Sumer, O., & Bader, G. D. (2015). Cytoscape. js: a graph theory library for visualization and analysis. Bioinformatics, 32(2), 309-311.

Huerta-Cepas, J., Serra, F., & Bork, P. (2016). ETE 3: Reconstruction, analysis, and visualization of phylogenomic data. Molecular biology and evolution, 33(6), 1635-1638.

Jakicic, J. M. (2002 Dec). The role of physical activity in the prevention and treatment of body weight gain in adults. Journal of Nutrition, 132(12): 3826S-3829S.

Jones, N., Furlanetto, D. L., Jackson, J. A., & Kinn, S. (2007). An investigation of obese adults’ views of the outcomes of dietary treatment. J Hum Nutr Diet., 20: 486–494.

Kazdin, A. E. (1974). Reactive self-monitoring: The effects of response desirability, goal setting, and feedback. Journal of Consulting and Clinical Psychology, 42(5), 704-716.

Kerren, A., Purchase, H. C., Ward, M., & Dagstuhl Seminar on Information Visualization– Multivariate Network Visualization. (2014). Multivariate network visualization: Dagstuhl Seminar #13201, Dagstuhl Castle, Germany, May 12-17, 2013 : revised discussions.

Liberati, A., Altman, D. G., Tetzlaff, J., Mulrow, C., Gøtzsche, P. C., Ioannidis, J. P., . . . Devereaux, P. J. (2009). The PRISMA Statement for Reporting Systematic Reviews and Meta-Analyses of Studies That Evaluate Health Care Interventions: Explanation and Elaboration.

Linde, J. A., Jeffery, R. W., French, S. A., Pronk, N. P., & Boyle, R. (2005 Dec). Self-weighing in weight gain prevention and weight loss trials. Annals of Behavioral Medicine, 30(3): 210-6.

Morton, K., Balazinska, M., Grossman, D., & Mackinlay, J. (2014). Support the data enthusiast: Challenges for next-generation data-analysis systems. Proceedings of the VLDB Endowment, 7(6), 453-456.

Nair, L., Shetty, S., & Shetty, S. (2016). Interactive visual analytics on Big Data: Tableau vs D3. js. Journal of e-Learning and Knowledge Society, 12(4).

Rohrer, J. E., Cassidy, H. D., Dressel, D., & Cramer, B. (2008). The effectiveness of a structured intensive weight loss program using health educators. Disease Management and Health Outcomes., 16, 449-454.

Sakaraida, T. J. (2010). Health promotion model. Mosby Elsevier.

Singh, A., Dey, N., Ashour, A., & Santhi, V. (2017). Web semantics for textual and visual information retrieval.

Tomar, G. (2017). The human element of big data: Issues, analytics, and performance.

U.S. Department of Health & Human Services. (2016, June 16). Defining Adult Overweight and Obesity. Retrieved from Centers for Disease Control and Prevention: https://www.cdc.gov/obesity/adult/defining.html

United States Preventative Services Taskforce. (2012). Screening for and management of obesity in adults. U.S. preventive services task force recommendation statement.

US Department of Agriculture and U.S. Department of Health and Human Services. (2010. Dec ). Dietary Guidelines for Americans. 7th Edition. Washington, DC: U.S. Government Printing Office.

Varona-Marin, D., Scott, S. D., & University of Waterloo. (2016). The lifecycle of a whiteboard photo: Post-meeting usage of whiteboard content captured with mobile devices.

Webber, K. H., Tate, D. F., & Quintiliani, L. M. (2008). Motivational interviewing in internet groups: a pilot study for weight loss. J Am Diet Assoc.., 108: 1029–1032.

Yon, B. A., Johnson, R. K., Harvey-Berino, J., Gold, B. C., & Howard, A. B. (2007). Personal digital assistants are comparable to traditional diaries for dietary self-monitoring during a weight loss program. Journal of Behavioral Medicine, 30: 165–175.

 

Bibliography

Automatic Detection of Dining Plates for Image-Based Dietary Evaluation: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3739713/

Dish Detection and Segmentation for Dietary Assessment on Mobile Phones: http://madima.org/madima2015/wpcontent/uploads/2015/10/Madima15_segmentation_final2.pdf

Food Image Analysis: Segmentation, Identification, and Weight Estimation: https://ieeexplore.ieee.org/document/6607548/

Lightweight Visual Data Analysis on Mobile Devices: Providing Self-Monitoring Feedback: https://kops.uni-konstanz.de/handle/123456789/39362

Mixed Reality Environments as Ecologies for Cross-Device Interaction: https://kops.uni-konstanz.de/handle/123456789/43028?locale-attribute=en

P.D. Leedy and Jeanne E. Ormond. “Practical Research: Planning and Design”. Pearson
Education, 2005. (Main library: FOLIO–001.42-LEE)

Rugg, G. “A Gentle Guide to Research Methods”. Open University Press, 2007. (Main
library: 378.242-RUG).

Snakes assisted food image segmentation: https://ieeexplore.ieee.org/document/6343437/

Specular Highlight Removal for Image-Based Dietary Assessment: https://ieeexplore.ieee.org/document/6266421/

Strunk, W. and White, E.B., “The Elements of Style”, Allyn and Bacon, 4th Edition,
2000. (Main library: 808-STR).

Swetnam, D. “Writing your dissertation: how to plan, prepare and present your work
successfully”, Oxford University Press, 3rd Edition, 2000. (Main library: 378.242-SWE).

 

Appendices

Appendix 1. Chart from The Visual Organization: Data Visualization, Big Data, and the Quest for Better Decisions on page 33

 

Appendix 2. Figure from Visual Analytics book by Daniel A. Keim

Appendix 3. Figure on building blocks of visual analytics

 

Appendix 4. Figures from Handbook of Data Visualization