Multi-Granular Text Encoding for Self-Explaining Categorization

Abstract

Self-explaining text categorization requires a classifier to make a prediction along with supporting evidence. A popular type of evidence is sub-sequences extracted from the input text which are sufficient for the classifier to make the prediction. In this work, we define multigranular ngrams as basic units for explanation, and organize all ngrams into a hierarchical structure, so that shorter ngrams can be reused while computing longer ngrams. We leverage a tree-structured LSTM to learn a contextindependent representation for each unit via parameter sharing. Experiments on medical disease classification show that our model is more accurate, efficient and compact than BiL-STM and CNN baselines. More importantly, our model can extract intuitive multi-granular evidence to support its predictions.