Long story short: finding health advice with informative summaries on health social media

Abstract

Purpose Whether automatically generated summaries of health social media can aid users in managing their diseases appropriately is an important question. The purpose of this paper is to introduce a novel text summarization approach for acquiring the most informative summaries from online patient posts accurately and effectively. Design/methodology/approach The data set regarding diabetes and HIV posts was, respectively, collected from two online disease forums. The proposed summarizer is based on the graph-based method to generate summaries by considering social network features, text sentiment and sentence features. Representative health-related summaries were identified and summarization performance as well as user judgments were analyzed. Findings The findings show that awarding sentences without using all the incorporating features decreases summarization performance compared with the classic summarization method and comparison approaches. The proposed summarizer significantly outperformed the comparison baseline. Originality/value This study contributes to the literature on health knowledge management by analyzing patients’ experiences and opinions through the health summarization model. The research additionally develops a new mindset to design abstractive summarization weighting schemes from the health user-generated content.

Identifying features of health misinformation on social media sites: an exploratory analysis

“…As shown in Figure 2, the feature list can be fed into several machine learning models to achieve model training, which includes decision tree, SVM, k-nearest neighbor classification (kNN), AdaBoost, gradient boosting decision tree (GBDT) and random forest. The literature has shown that the identification of health misinformation on SMSs is mainly based on surface features such as punctuation marks, pictures and symbols, but rarely on the semantic features and source features of health misinformation (Liu et al , 2019; Søe, 2018; Zhao et al , 2021). Therefore, this study finds an approach to identify health misinformation on SMSs through semantic features and source features.…”

“…As this study explores what are the features of health misinformation on SMSs – a theme that has received little research attention – and proposes relatively exploratory research questions, an inductive approach was best suited for coding and analysis (Li et al , 2018; Liu et al , 2019). Specifically, due to the limited literature, it is impossible to develop a set of priority codes, so open coding, axial coding and selective coding are used to code the features of health misinformation (Corbin and Strauss, 1990; Miles et al , 2013).…”

“…For example, Ruchansky et al (2017) proposed the CSI (capture, score and integrate) model that identifies three features of fake news on websites: article text, user response and source user. Liu et al (2019) extracted the features of health rumors on the internet into three categories: writing style, text topic and others.…”

Social Media, Grindr, and PrEP: Sexual Health Literacy for Men Who Have Sex with Men in the Internet Age

“…The development of novel information technology tools in order to support online HIV consumer health is another research path holding great potential. Current research in this area includes: the development of an ontology-based FAQ query system for HIV/AIDS information [39], future versions of which could potentially be incorporated into GSN app sexual health information sections; continued research into social media data mining and monitoring to help make informed decisions and strategies concerning online HIV information materials [20,27,28]; and the creation of automated text summaries from health social media posts [40].…”

Personalized content recommendation in online health communities

“…However, there are several factors that impede the social support exchange in OHCs. First, content in these platforms are vast and heterogeneous and users are always easily overwhelmed by the ocean of information which leads to the information overload problem (Liu et al, 2019b). Community members tend to participate and remain engaged in the discussions that are perceived relevant and beneficial and tailored content are more likely to be read, understood, accepted and perceived as credible compared to one-size-fits-all content (Wang et al, 2019).…”