Using Text Mining Tools to Identify Preference and Context Information from User’s Opinion

SILVANA ACIAR

PERSONAL AND UBIQUITOUS COMPUTING

SPRINGER LONDON LTD

Año: 2011 p. 1 - 8

1617-4909

Consumer reviews, opinions and shared experiences are popularways to express preferences and interest of tourists in most of thepopular tourism sites inside of the typical valuation of product using anyvaluation scale. The most critical issue on opinion mining is how toextract information that can be understood and utilized by computers fromwritten text by users/consumers in natural language. Several approachesusing artificial intelligence have been used to deal with this problemeven so; problem that has been less addressed but not less important isthe identification of context information embedded in consumer’s opinionsdue the arduous task of processing natural language in which reviews havebeen expressed. This paper addresses this problem based on classificationtext mining techniques to identify review’s sentences containingcontextual information to be then processing and incorporated in arecommender system. This approach was exemplified by a case study usingreviews from www.tripadvisor.com.