Sentiment Mining of Product Opinion Data
Vikas Thada1, Utpal Shrivastava2, Ruchi3
1Dr. Vikas Thada*, Department of Computer Science & Engineering Amity University Gurgaon
2Mr Utpal Shrivastava, Department of Computer Science & Engineering Amity University Gurgaon
3Ms Ruchi, Department of Computer Science & Engineering Amity University Gurgaon
Manuscript received on December 16, 2019. | Revised Manuscript received on December 28, 2019. | Manuscript published on January 10, 2020. | PP: 1218-1222 | Volume-9 Issue-3, January 2020. | Retrieval Number: C8641019320/2020©BEIESP | DOI: 10.35940/ijitee.C8641.019320
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Sentiment Analysis has gotten a focal point of consideration for the retrieving of web information and text corpora. At the point when we talk about sentiment analysis, it investigates the people’s criticism or conclusion or audits in the wake of removing those reviews from various stages. While, when we find out about Sentiment Analysis, it recognizes the mining or analysis, which are communicated through content and afterward we investigate it. The fundamental goal of Sentiment Analysis is to reason feelings and ideas. In this paper the research work has done sentiment analysis on Amazon Reviews. The data set used is Amazon Reviews on Unlocked Mobile phones dataset of 413840 records. There are number of columns for the dataset like Product Name, Price, Rating, Brand Review text and the count of people for whom the review was helpful. But the research work has used on the Rating and Reviews columns. The research work has applied bag of words, tf-idf and n-grams techniques. It was found that n-gram with bag of words approach gave a maximum testing accuracy of 97% and AUCROC score of .967. Using deep learning approaches and their comparing with used ones is left as future work.
Keywords: Sentiment, Analysis, Opinion, Mining, Review, bag of words, N-gram.
Scope of the Article: Data Mining