Marketing decision making using sentiment score with co-extracting opinion target words from reviews. Issue 1 (February 2021)

• Record Type: Journal Article
• Authors: Saravanan, T M; Kavitha, T; Hemalatha, S; Kumar, Gourav
• Is Part Of: IOP conference series. Volume 1055:Issue 1(2021)
• Publication Date: 2021-02
• Subjects: Materials science -- Periodicals; 620.1105
• DOI: 10.1088/1757-899X/1055/1/012074
• Languages: English
• ISSNs: 1757-8981

Abstract

Abstract: This paper proposes a broad-spectrum outline that utilizes natural language processing approaches counting sentiment analysis, web mining that uses clustering schemes to acquire fresh outcome elicited from reviews in support of diverse features of products. An ultimate part in the scheme is finding positive and negative sentiment scope for all features of a product in the given reviews. Massive figures of product reviews are crawling on the Web with the rapid development of Internet. Commencing these reviews, customers can get hold ofactualexamination of product rank and unswerving supervision of their acquisition actions. Withdrawal feelings from web reviews have developed into a gradually more fiery activity and haveengrossed a vastpact of considerationcommencingthe researchers. The key task for opinion mining is obtaining opinion emotions from reviews in online, the novelfactor involves in identifying relations between words. The project proposes a new scheme that contains graph based co-ranking scheme that is employed in finding poise of every word. Then, words with superiorpoise are withdrawn as opinion objects. Based on the nearest neighbor rule we compare with previous schemes, the scheme more preciselycaptures opinion relations, particularly for elongated cover relations. The investigationaloutcome shows that the schemesuccessfully outperforms all other methods and techniques.