Home // SOTICS 2022, The Twelfth International Conference on Social Media Technologies, Communication, and Informatics // View article

#MyIBDHistory on Twitter: Predicting IBD Type from Personal Tweets

Authors:
Maya Stemmer
Gilad Ravid
Yisrael Parmet

Keywords: Twitter; IBD; data analysis; logistic regression.

Abstract:
Inflammatory Bowel Disease (IBD) is a chronic inflammation condition of the digestive system that is usually classified into one of two diseases: Crohn's Disease (CD) or Ulcerative Colitis (UC). If neither one is diagnosed with certainty, the patient is diagnosed with IBD Unclassified (IBD-U). IBD patients form communities on Twitter and exchange thoughts regarding their diseases. In 2018, IBD patients shared their disease history on Twitter and signed their tweets with the hashtag #MyIBDHistory. In their tweets, they mentioned their age at diagnosis, the medications they have tried over the years, whether they underwent any surgeries, and more. In this research, we analyzed patients' tweets containing the hashtag #MyIBDHistory and built a classifier that predicts the IBD type (CD or UC) of a patient. We transformed the disease history described in the tweets into tabular classification features and assessed their importance. We identified key features that helped distinguish CD from UC and used the classifier to predict the disease type of IBD-U patients. Our results were correlated with IBD related research, as the two most prominent features that tilted the classification towards CD were having a fistula and suffering from nutrient deficiency.

Pages: 13 to 19

Copyright: Copyright (c) IARIA, 2022

Publication date: October 16, 2022

Published in: conference

ISSN: 2326-9294

ISBN: 978-1-68558-002-5

Location: Lisbon, Portugal

Dates: from October 16, 2022 to October 20, 2022