CAP6640 Natural Language Processing

Vector Semantics and NLP Applications:

Search Engines, Business Intelligence, and Cybersecurity

By: Dr. Wingyan Chung

Objectives

Upon successfully completing this module, students should be able to:

1. Describe vector models and their applications in natural language processing (NLP).
2. Explain concepts in term weighting, PPMI, inverted index, and measures of similarity in vector models.
3. Apply deep learning to word and context prediction using skip-gram and CBOW models.
4. Evaluate quality of prediction outcomes on word and context.

Readings

1. News: "Obama wants you to join CyberCorps Reserve"
2. News: White House wants to revamp cybersecurity
3. Memex (Domain-Specific Search), DARPA I2O
4. Text: "Vector Semantics," in Jurafsky & Martin "Speech and Language Processing (3rd ed. draft)"
5. Slides on Vector Semantics
6. "Deep or Shallow, NLP is Breaking Out," Communications of the ACM, March 2016

NLP Research and Applications

1. Word2Vec: Deep Learning of Text Corpus
2. IPM Special Issue: Emotion and Sentiment in Social and Expressive Media
3. Contextual semantics for sentiment analysis of Twitter
4. Sentiment and network analyses of U.S. Immigration and border security
5. BizPro: Extracting and categorizing business intelligence factors from textual news articles
6. Text Visualization for Authorship Analysis
7. Web Searching in a Multilingual World, CACM

Copyright © 2016 Wingyan Chung. All Rights Reserved.