KEY TERMS AND DEFINITIONS

Knowledge Discovery:

Research field in computer science concerned with devising processes and computational methods aiming at automatically detecting novel, useful patterns in very large volumes of data where the use of traditional data analysis techniques relying on human inspection alone are not feasible or too costly.

Opinion Lexicon:

A lexical resource in a given language, typically in the form of a machine readable database that categorizes terms according to their subjective content and opinion polarity.

Opinion Mining:

Research area concerned with the application of computational methods to the treatment of subjectivity in text. Opinion mining is a cross disciplinary area that leverages and contributes to advances in text mining, natural language processing and machine learning.

Sentiment Classification:

In opinion mining, refers to the task of predicting if a given piece of text, known a priori to be subjective, conveys positive, negative or neutral sentiment by applying automatic computational methods derived from text mining approaches, or by leveraging natural language processing techniques.

SentiWordNet:

An opinion lexicon for the English language built upon the WordNet database of terms and relationships. SentiWordNet is built by expanding from a core of terms known to carry a priori sentiment information (e.g. "good", "bad", etc.) by using WordNet's term relationships and each term's explanatory text, or glosses.

Subjectivity Detection:

In opinion mining, refers to application of predictive methods to determine the degree of which a given piece of text is subjective.

Subjectivity:

In text, refers to the use of language for the expression of private states of mind such as opinions, judgement, evaluations and speculations.

WordNet:

WordNet is a database of terms and relationships for the English language, where words are grouped into sets of similar meaning (synsets), and may contain semantic associations such as synonym, antonym, "is-a", etc.