Text sentiment analysis on E‐shopping product reviews using chaotic coyote optimized deep belief network approach. (7th May 2022)

• Record Type: Journal Article
• Authors: Mohana, R. S.; Rajathi, K.; Kousalya, K.; Yuvaraja, T.
• Is Part Of: Concurrency and computation. Volume 34:Number 19(2022)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2022-05-07
• Subjects: chaotic coyote optimization algorithm -- deep belief network -- E‐shopping reviews -- text sentiment analysis; Parallel processing (Electronic computers) -- Periodicals; Parallel computers -- Periodicals; 004.35
• DOI: 10.1002/cpe.7039
• Languages: English
• ISSNs: 1532-0626

Abstract

Abstract: Text sentiment analysis is mainly used to the customers benefits. In the existing works, the text sentiment analysis faces more troubles such as, disambiguation (removing unwanted terms), discussions, contrast, intensity, and excessive flections and complex sound structure with less accuracy. In this article, the text sentiment analysis on E‐shopping product using chaotic coyote optimized deep belief network approach is proposed to minimize the troubles in the sentiment analysis and increase the accuracy. The other name of sentiment analysis is subjective analysis. The main objective of this article is "classify the text sentiments according to the polarity (positive and negative) and to increase the accuracy." Here, the E‐shopping Kaggle datasets are preprocessed and the features are extracted. Then, the extracted features of the trained data's are given using deep belief network (DBN) classifier to get pure sentiments (positive or negative) with accuracy. Here, the performance metrics of the accuracy, recall, and precision, F‐measure, specificity, and sensitivity are calculated. The simulation process is executed in Python platform. The proposed chaotic coyote optimized deep belief network (CCO‐DBN) attains accuracy 9.8%, precision 17.2%, recall 5.61%, F‐measure 17.07%, specificity 2.247%, sensitivity 13.25% is higher than the existing methods such as FCM‐DFA, GA, SVM‐RFA.