Downloads: 107 | Views: 229
Survey Paper | Computer Science & Engineering | India | Volume 3 Issue 11, November 2014 | Rating: 7 / 10
A Survey : Extracting Features from Online Reviews Corpus by Domain Relevance
Ashwini P. Lingojwar | L. J. Sankpal [5]
Abstract: Large amount of user generated data is present on web in the form of blogs, reviews tweets, comments etc. This type of data involves users opinion, view, attitude, sentiment towards particular product, topic, event, news etc. Capturing public opinion about social events and product preferences is increasing interest from the customer and from the business world. Nowadays, if one wants to by a product, consumer is no longer limited to asking friends and family for opinions because there are many user reviews and discussions in public forums on the Web about the product. But consumers are not satisfied with overall reviews of the product where as they want to understand which is positive and negative attribute of the product. Opinions and sentiments are expressed in the text reviews. Most of the opinion features are extracted from the online review corpus by using mining patterns. As Features are extracted from single review corpus it ignoring the important difference in word distributional characteristics of opinion features across different corpora. As a result efficient opinion features are not extract. In order to obtain valid feature it is necessary to consider two different online review according to the domain relevance.
Keywords: Opinion mining, natural language processing, opinion feature, information search and retrieval
Edition: Volume 3 Issue 11, November 2014,
Pages: 2303 - 2305