11:00am - 11:10amID: 163
/ PS-09: 1
Topics: Data Science; Analytics; and VisualizationKeywords: systematic reviews, health controversies, public health, evidence synthesis, network visualization
Visualizing Evidence-Based Disagreement Over Time: The Landscape of a Public Health Controversy 2002-2014
Systematic reviews answer specific questions based on primary literature. However, systematic reviews on the same topic frequently disagree, yet there are no approaches for understanding why at a glance. Our goal is to provide a visual summary that could be useful to researchers, policy makers, and health care professionals in understanding why health controversies persist in the expert literature over time. We present a case study of a single controversy in public health, around the question: “Is reducing dietary salt beneficial at a population level?” We define and visualize three new constructs: the overall evidence base, which consists of the evidence summarized by systematic reviews (the inclusion network) and the unused evidence (isolated nodes). Our network visualization shows at a glance what evidence has been synthesized by each systematic review. Visualizing the temporal evolution of the network captures two key moments when new scientific opinions emerged, both associated with a turn to new sets of evidence that had little to no overlap with previously reviewed evidence. Limited overlap between the evidence reviewed was also found for systematic reviews published in the same year. Future work will focus on understanding the reasons for limited overlap and automating this methodology for medical literature databases.
11:10am - 11:20amID: 216
/ PS-09: 2
Topics: Social Media and Social ComputingKeywords: Diet, Social Media, Twitter, Text Mining, Topic Modeling
Seasonal Characterization of Diet Discussions on Reddit
1University of South Carolina, USA; 2Johns Hopkins University, USA
To monitor public opinions on diet, large survey data is commonly used though costly and time consuming. Social media has become a mainstream channel of communication and has drastically grown in popularity, connecting users to a ready stream of health information. While the literature has provided valuable information on the use of social media for health, a broader perspective informed by different types of social media platforms would be highly beneficial. Diet has been extensively explored on a few mainstream platforms, further informing public health research. However, diet conversations on Reddit have only been studied within a narrow scope, looking at specific sub-communities. This study aims to characterize diet-related posts and their seasonal patterns using a mixed method approach. We collected more than 500,000 posts with subsequent comments from Reddit over the course of a year. Our findings show that Reddit users across all sub-communities primarily discussed health promotion, fitness plans, a healthy lifestyle, diet and fitness progress, food experiences, weight loss, as well as vegan and vegetarian diets. In addition, seasonal differences based on the weight of most topics, were found to be significant (p < 0.05).
11:20am - 11:30amID: 276
/ PS-09: 3
Topics: Human Computer Interaction (HCI)Keywords: Acceptability, artificial intelligence, trust, diagnostic results, healthcare.
Lay Individuals’ Perceptions of Artificial Intelligence (AI)-Empowered Healthcare Systems
1Pace University, USA; 2Florida State University, USA; 3IBM Research, USA; 4Chinese Academy of Sciences, China
With the recent advances in Artificial Intelligence (AI) technology, more and more patient-facing applications have started embodying this novel technology to deliver timely healthcare information and services to the patient. However, little is known about lay individuals’ perceptions and acceptance of AI-driven, patient-facing health systems. In this study, we conducted a survey with 203 participants to investigate their perceptions about using AI to consult information related to their diagnostic results and what factors influence their perceptions. Our results showed that despite the awareness and experience of patient-facing AI systems being low amongst our participants, people had a generally positive attitude towards such systems. Several intrinsic factors, such as education background and technology literacy, play an important role in people’s perceptions of using AI to comprehend diagnostic results. We conclude this paper by discussing the implications of this work, with an emphasis on enhancing the trustworthiness of AI and bridging the digital divide.
11:30am - 11:40amID: 306
/ PS-09: 4
Topics: Data Science; Analytics; and VisualizationKeywords: healthy diet recommendation, graph mining, data integration
Healthy Diet Recommendation via Food-Nutrition-Recipe Graph Mining
1Dalian Maritime University, People's Republic of China; 2Zhejiang University, Hangzhou, People's Republic of China; 3Alibaba Group, People's Republic of China; 4Indiana University Bloomington, USA
Good nutrition and balance dietary pattern play vital roles of leading a healthy lifestyle. Prior studies showed that the healthy diet can successfully reduce the risk of chronic diseases (e.g., type 2 diabetes and cancer) and bring other well-documented benefits. Existing food recommendation models, however, often solely rely on user's feedback (e.g., click and purchase data), which aims to optimize Click-Through Rate (CTR) but ignores the importance of health needs of users. Intuitively, a healthy diet recommendation requires a comprehensive consideration of different kinds of information, such as nutrition, ingredients and cooking methods. In this study, by collecting the data from FoodData Central (FDC), recipe websites, and scientific literature, we construct a heterogeneous graph, Food-Nutrition-Recipe Graph (FNRG), by integrating information of nutrition, food (ingredients), and recipes. A random walk based graph mining approach is proposed to meet the health needs of users. Experiments results show that the proposed method can successfully address the health information needs for people who suffer from chronic diabetes.
11:40am - 11:50amID: 351
/ PS-09: 5
Topics: Domain-Specific InformaticsKeywords: eHealth Literacy, Information Source, Reading Behavior, Eye-tracking, Online Health Information Seeking
EHealth Literacy, Information Sources, and Health Webpage Reading Patterns
The University of Texas at Austin, United States of America
A lab-based experiment was conducted to understand how eHealth literacy and information source affect reading vs. scanning behavior on health webpages. Participants read 15 webpages from commercial, government, and online forum sources while their eye movements were tracked. Negative binomial regression and Kruskal-Wallis tests revealed that high eHealth literacy participants tend to scan webpages, while low literacy participants tend to read webpages. There were no differences in the tendencies to scan or read among different information sources. Our work shows that observable objective information behavior is attributable to eHealth literacy and may provide additional insights to the measurement of eHealth literacy.