Dobór optymalnego zestawu słów istotnych w opiniach konsumentów na potrzeby ich automatycznej analizy
Selection of the optimal set of relevant words in consumers opinions in the context of the opinion mining
Author(s): Janusz Tuchowski, Katarzyna WójcikSubject(s): Economy
Published by: Wydawnictwo Uniwersytetu Ekonomicznego we Wrocławiu
Keywords: text-mining; Web-mining; taxonomy; text document classification; evaluation of measures
Summary/Abstract: Sentiment analysis is the research area that can have a significant impact on to-day’s business. A lot of sentiment analysis studies base on similarity (or distance) between opinions as special kinds of texts. The problem that must be solved is to determine the best similarity measure to deal with opinions. The main objective of this paper is to conduct the analysis of the selection of optimal set of words relevant when calculating the similarity (distance) of text documents in the context of the opinion mining. Computations are con-ducted in R language, RapidMiner application and spreadsheet.
Journal: Prace Naukowe Uniwersytetu Ekonomicznego we Wrocławiu
- Issue Year: 2014
- Issue No: 328
- Page Range: 106-115
- Page Count: 10
- Language: Polish