JMP Pro 16 uses Sentiment Analysis to score freeform text from logs, journals, reports or surveys based on a dictionary of terms that have associated scores. This gives you an overall sense of how positive or negative someone feels about a product, process or situation. Some words are considered as negaters and intensifiers to assign the proper sentiment. For example, the phrases “I like ice cream”, “I really like ice cream”, “I don’t like ice cream” and “I really don’t like ice cream” result in different sentiments.
JMP Pro 16 Term Selection uses supervised machine learning techniques to identify the relevant terms and assign to them a score that you would possibly use in a sentiment analysis to help you build a predictive model. For example, suppose you have \product reviews on a product and the star rating assigned. Term Selection will allow you to determine key terms of interest in predicting an outcome, high (and low) star ratings.
Resources