TopicClassifiers folder contains code to identify themes in reviews, assign tags to them, and develop a classifier based on that. Unsupervised + supervised approaches.
Clustering Books IPyNb contains code to cluster the Gutenberg and Brown corpora into genres and then use similarity functions to recommend relevant books from that corpus based on a query. Query is a book title whose plot is scraped from Wikipedia to retrieve content for matching.
CORELab Evaluation folder contains code for text clustering / summarization of short answer questions in introductory CS examinations to automatically group answers into score buckets.