TODOs - LDAforVisProject/LDA GitHub Wiki

To do list:

• Learning Contract (proposal finished)
• Consider singular occurrence --> higher order topic mapping
• Optional: Additional text-preprocessing methods (stemming, lemmatization, merging of words such as "multi-variate" and "multivariate")
• Replacing cosine distance similarity with rescaled dot product
• Manually extract about 20 abstracts and run through algorithm (done)
• Replacement of multiword technical terms with a single term using underscores (done)