Computational Approaches Toward Integrating Quantified Self Sensing and Social Media

Abstract

The growing amount of data collected by quantified self tools and social media hold great potential for applications in personalized medicine. Whereas the first includes health-related physiological signals, the latter provides insights into a user’s behavior. However, the two sources of data have largely been studied in isolation. We analyze public data from users who have chosen to connect their MyFitnessPal and Twitter accounts. We show that a user’s diet compliance success, measured via their self-logged food diaries, can be predicted using features derived from social media: linguistic, activity, and social capital. We find that users with more positive affect and a larger social network are more successful in succeeding in their dietary goals. Using a Granger causality methodology, we also show that social media can help predict daily changes in diet compliance success or failure with an accuracy of 77%, that improves over baseline techniques by 17%. We discuss the implications of our work in the design of improved health interventions for behavior change.

Characterizing and Predicting Repeat Food Consumption Behavior for Just-in-Time Interventions

“…Next, recent studies have investigated the healthiness cues uncovered from food images posted by Instagram users [24] and online cooking recipes from Allrecipes [27]. Lastly, online food diaries data from MyFitnessPal users have been used to study individuals' dieting [8,16,34], food substitutes extraction [3], and healthy eating behaviors [2]. In contrast to the public health monitoring aspect of previous work, our work focuses on predicting food items likely to be consumed in the next consumption session which has a direct application to the just-in-time health interventions.…”

Does Journaling Encourage Healthier Choices?

“…Mejova et al [31] analyzed food pictures shared by Instagram users to study the prevalence of obesity. Recently, a few studies have investigated the tasks of predicting diet compliance outcomes using MFP food diary data [46] together with Twitter data [17].…”

“…Our work is highly relevant to [17,46] in which the researchers constructed computational models to predict diet compliance success using different types of features, such as words & food types identified from MFP diary entries and social and linguistic attributes extracted from the users' social media messages. While their studies particularly focused on caloric balance as the primary outcome, we examine a more comprehensive set of eating behaviors by using evidence-based healthy eating outcomes as the primary measures.…”

“…For some, this may positively indicate that they have already achieved their self-tracking goals -an example of successful abandonment [12]. In a recent study, De Choudhury et al [17] concluded that this phenomenon should not be common and long-term food journalers are more likely to complete their diary entries when they choose to record a journal.…”

Eat & Tell

“…Another related body of work focuses on analyzing self-tracking data in conjunction with other data sources, e.g., social media messages, to learn about individuals' health-related behaviors. For example, De Choudhury et al [9] proposed computational methods to predict individuals' diet compliance success using linguistic, activity, and social capital features extracted from their Twitter messages. Wang et al [35] studied weight updates automatically shared on Twitter from a Withings smart scale.…”