Sentiment Analysis

Sentiment analysis refers to the evaluation of the sentiment – typically positive or negative – of the text based on how the language is used. It is about determining the general tone of a document based on the application of computational linguistics algorithms. A typical goal might be to measure consumers’ reactions to a new marketing campaign and to correlate the findings to the expected financial impact on the business.

FAST ESP’s Sentiment Analysis determines the tone of a document by breaking the document into its basic parts of speech. This parts-of-speech tagging identifies all the structural elements of a document or sentence, including verbs, nouns, adjectives, proper nouns, etc. The feature identifies the parts of speech, which indicate emotion. In most cases they are adjective-noun combinations such as “horrible sight” or “devastating loss.” Then the tool scores the phrases according to their tone.

The tone of a document is evaluated by determining how frequently a given phrase occurs near a set of “good” words (e.g., good, wonderful and spectacular) and in proximity to a set of “bad” words (e.g., bad, horrible and awful). The Sentiment Analysis feature computes how close the good and bad words are to the phrase under consideration.

Source: FAST ESP Brochure 2007