CS 572 - Information Retrieval final project, Emory University 2016

• Semantic Text Similarity Dataset Hub
• Trec LiveQA 2016
• Java Skeleton Server

• L6 from Yahoo Webscope
• Wikipedia Dump
• Reddit Comment Dump (need to figure out how to link comments with posts)

• Serialize documents in Yahoo QA XML files as an SQLite database
• Build HTTP server for recieving questions for the competition: See Java example on webpage
• Make interface for adding generic prediction models
• Write mixture of experts predictor given all prediction models models
• Article summarizer (use on Wikipedia, Reddit to add more recent data)

• TF-IDF or BM-25 - Gensim
• Latent Semantic Indexing - Gensim
• Latent Dirichlet Allocation - Gensim
• Word2Vec - Gensim

• Neural network model - Some success using Keras