Perspective Discovery in Controversial Debates
Date created: | Last Updated:
: DOI | ARK
Creating DOI. Please wait...
Description: This repository contains all the material used for the research on Perspective Discovery in Controversial Debates. We provide the code, data and metadata in order to reproduce the research done. We researched the potential of topic models to automatically discover perspectives in text. For this we evaluated five topic models (LDA, TAM, VODUM, JST and LAM) in terms of their clustering ability of perspective labels and human understandability of the model's output. The results show that some of these models may discover perspectives from text.