Supporting Personalized Pregnancy Care Using Artificial Intelligence: Identifying Online Health Education Materials to Prevent Postpartum Depression Using Natural Language Processing
Model 1: This model is trained to classify articles that are relevant and irrelevant to pregnancy.
Tag "1": Related Articles (These are articles related to Pregnancy)
Tag "4": Unrelated Articles (These are articles that are unrelated to Pregnancy)
Model 2: This model is trained to categorize articles related to Pregnancy: Diet, Exercise, Mental Health, or other (Articles that do not fall into these 3 categories)
Tag "1": Articles related to Diet (Pregnancy-Related)
Tag "0": Articles related to Exercise (Pregnancy-Related)
Tag "3": Articles related to Mental Health (Pregnancy-Related)
Tag "4": Articles related to the broader topic of Pregnancy itself but not related to the above 3 main categories
One-step: This model is trained to identify articles into 5 categories in a single model: Diet, Exercise, Mental Health, Other, or Unrelated
Tag "1": Articles related to Diet (Pregnancy-Related)
Tag "0": Articles related to Exercise (Pregnancy-Related)
Tag "3": Articles related to Mental Health (Pregnancy-Related)
Tag "4": Articles related to the broader topic of Pregnancy itself but not related to the above 3 main categories
Tag "9": Unrelated Articles (These are articles that are unrelated to Pregnancy)
Overall or Two-step: This model is a combination of Model 1 and Model 2.
Contact: Zhaoyi Sun (zhs4003[at]med[dot]cornell[dot]edu) or Braja Gopal Patra (brajagopal[dot]cse[at]gmail[dot]com)