CS 670 - Natural Language Processing

Prerequisites, admission to a graduate computational and data sciences program and CS 501 or CS 510 or consent of instructor. Natural Languages have evolved from thousands of years of human existence as they pass from generation to generation. The grammar of any natural language is complex and different from other languages. Moreover, it is evolutionary. This makes Natural Language Processing (NLP) a complex challenge. The main goal of NLP is to understand the meaning of text. Only when computers understand the real meaning of the text, can they take decisive action which must be the intended action. Sentiment analysis of text is one of the important applications of NLP. The use case of sentiment analysis is for the purpose of analyzing customer feedback and tweets. The translation of text between languages is another significant NLP application. NLP capabilities are currently used in daily lives. Personal assistant software like Siri (Apple’s iPhone), Google Assistant (search engines), Amazon’s Alexa, and robots use NLP (text and audio) to understand the user’s intent and provide an accurate response. Spelling and grammar errors flagged by word processing software packages (like MS Word, Google Docs) use NLP to improve the written text. Mobile phones use NLP to understand and transcribe audio messages. There are currently two different approaches to NLP. The first one is the analysis of words, sentences, and the semantics of text. There are various software packages that can provide these capabilities. These software packages are Natural Language Tool Kit (NLTK), TextBlob, and spaCy. The other approach to NLP is using the Machine Learning strategy to analyze the text. Neural Network models are used to train a model by feeding it a lot of text data. Google Cloud Platform (GCP) provides a Machine Learning (ML) API (Application Programming Interface) for the analysis of Natural Languages and provides translation service between languages. IBM Watson provides similar services for NLP. Letter grade. (Offered every year.) 3 credits