Clustering and Topic Modeling in Python

This workshop addresses clustering and topic modeling in Python, primarily through the use of scikit-learn and gensim. We first read in a corpus, prepare the data, create a tfidf matrix, and cluster using k-means. We will then compare results to LSI and LDA topic modeling approaches.

Prerequisites: Attendees should either already have a thorough knowledge of Python, or have attended the Python for Everything series.

Please install the following packages ahead of the workshop:

Python 3 (https://www.continuum.io/downloads)

Packages:

• NLTK ( $ pip install nltk) 
• scikit-learn ( $ pip install scikit-learn) 
• pandas ( $ pip install pandas) 
• matplotlib ( $ pip install matplotlib) 
• gensim ( $ pip install gensim)

Dataset: http://www.cs.cmu.edu/~dbamman/booksummaries.html