Abstract

RevieweR is a visualization tool that helps in interpretation of topic models and seeing the change in perspectives on a topic through text analysis of literature on that subject.

Topic modeling is used to study a set of abstracts and find clustering of papers or main themes that can be gleaned.

This tool helps in checking which topics are being selected and how this choice varies with the hyperparameter of number of topics to be found. Also, to finetune the modeling process by checking how well the lemmatization, stop words and number of topics chosen is working.

Finally, once the topic is selected, we want to pull out the papers associated with that topic and infer the changes in the main themes studied in the literature.