2015 | OriginalPaper | Buchkapitel
Developing Term Weighting Scheme Based on Term Occurrence Ratio for Sentiment Analysis
verfasst von : Nivet Chirawichitchai
Erschienen in: Information Science and Applications
Verlag: Springer Berlin Heidelberg
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Term weighting is an important task for sentiment classification. Inverse document frequency (IDF) is one of the most popular methods for this task; however, in some situations, such as supervised learning for sentiment classification, it doesn’t weight terms properly, because it neglects the category information and assumes that a term that occurs in smaller set of documents should get a higher weight. In this paper, I purpose sentiment classification framework focusing on the comparison of various term weighting schemes, including Boolean, TF, TFIDF and a novel term weighting (TOW). I have evaluated these methods on Internet Movie Database corpus with four supervised learning classifiers. I found TOW weighting most effective in our experiments with SVM NB and NN algorithms. Based on our experiments, using TOW weighting with SVM algorithm yielded the best performance with the accuracy equaling 93.45 %.