Natural Language Processing
Extracted features from unstructured text data (bag of words, quality phrase mined (SegPhrase), CoreNLP, sentiment analysis, parts-of-speech (POS) tagging, tf-idf, topic mining, PLSA, Latent Dirichlet Allocation (LDA), and Non-Negative Matrix Factorization (NNMF). Predicted restaurant hygiene from restaurant reviews (Yelp) with 0.87 recall and 0.87 precision, created resume search engine from web scraped resumes, and developed recommender system for restaurant by popular dish from reviews.