%0 Journal Article
%@ 2291-9694
%I JMIR Publications
%V 10
%N 7
%P e37201
%T Extraction of Explicit and Implicit Cause-Effect Relationships in Patient-Reported Diabetes-Related Tweets From 2017 to 2021: Deep Learning Approach
%A Ahne,Adrian
%A Khetan,Vivek
%A Tannier,Xavier
%A Rizvi,Md Imbesat Hassan
%A Czernichow,Thomas
%A Orchard,Francisco
%A Bour,Charline
%A Fano,Andrew
%A Fagherazzi,Guy
%+ Center of Epidemiology and Population Health, Inserm, Hospital Gustave Roussy, Paris-Saclay University, 20 Rue du Dr Pinel, Villejuif, 94800, France, 33 142115386, adrian.ahne@protonmail.com
%K causality
%K deep learning
%K natural language processing
%K diabetes
%K social media
%K causal relation extraction
%K social media data
%K machine learning
%D 2022
%7 19.7.2022
%9 Original Paper
%J JMIR Med Inform
%G English
%X Background: Intervening in and preventing diabetes distress requires an understanding of its causes and, in particular, from a patient’s perspective. Social media data provide direct access to how patients see and understand their disease and consequently show the causes of diabetes distress. Objective: Leveraging machine learning methods, we aim to extract both explicit and implicit cause-effect relationships in patient-reported diabetes-related tweets and provide a methodology to better understand the opinions, feelings, and observations shared within the diabetes online community from a causality perspective. Methods: More than 30 million diabetes-related tweets in English were collected between April 2017 and January 2021. Deep learning and natural language processing methods were applied to focus on tweets with personal and emotional content. A cause-effect tweet data set was manually labeled and used to train (1) a fine-tuned BERTweet model to detect causal sentences containing a causal relation and (2) a conditional random field model with Bidirectional Encoder Representations from Transformers (BERT)-based features to extract possible cause-effect associations. Causes and effects were clustered in a semisupervised approach and visualized in an interactive cause-effect network. Results: Causal sentences were detected with a recall of 68% in an imbalanced data set. A conditional random field model with BERT-based features outperformed a fine-tuned BERT model for cause-effect detection with a macro recall of 68%. This led to 96,676 sentences with cause-effect relationships. “Diabetes” was identified as the central cluster followed by “death” and “insulin.” Insulin pricing–related causes were frequently associated with death. Conclusions: A novel methodology was developed to detect causal sentences and identify both explicit and implicit, single and multiword cause, and the corresponding effect, as expressed in diabetes-related tweets leveraging BERT-based architectures and visualized as cause-effect network. Extracting causal associations in real life, patient-reported outcomes in social media data provide a useful complementary source of information in diabetes research. 
%M 35852829
%R 10.2196/37201
%U https://medinform.jmir.org/2022/7/e37201
%U https://doi.org/10.2196/37201
%U http://www.ncbi.nlm.nih.gov/pubmed/35852829