Impact of Customer Feedback System using Machine Learning Algorithms for Sentiment Mining
B. Hemalatha1, T. Velmurugan2

1B.Hemalatha*, Assistant Professor1 Department of Computer Applications, Margregorios College, University of Madras, Chennai, Tamil Nadu, India.
2T. Velmurugan, PG and Research Department of Computer Science, D. G. Vaishnav College, University of Madras, Chennai. Tamil Nadu, India.
Manuscript received on January 10, 2020. | Revised Manuscript received on January 22, 2020. | Manuscript published on February 10, 2020. | PP: 1475-1483 | Volume-9 Issue-4, February 2020. | Retrieval Number: D1537029420/2020©BEIESP | DOI: 10.35940/ijitee.D1537.029420
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: The process of discovering and analyzing the customer feedback using Natural Language Processing (NLP) is said to be sentiment analysis. Based on the surge over the concept of rating level in sentiment analysis, sentiment is utilized as an attribute for certain aspects or features that get expressed and more attention are provided to the problem of detecting the customer reviews. Despite the wide use and popularity of some methods, a better technique for identifying the polarity of a text data is hard to find. Machine learning has recently attracted attention as an approach for sentiment analysis. This work extends the idea of evaluating the performance of various Machine Learning (ML) classifiers namely logistic regression, Naive Bayes, Support Vector Machine (SVM) and Neural Network (NN).To show their effectiveness in sentiment mining of customer product reviews, the customer feedback has been collected from Grocery and Gourmet Food. Nearly 90 thousands customers feedback reviews of various product related categories namely Product ID, rating, review test, review time reviewer ID and reviewer name are used in this analysis. The performance of the classifiers is measured in terms of accuracy, specificity and sensitivity. From the experimental results, the better machine learning classification algorithm is proposed for sentiment mining using online shopping customer review data.
Keywords: Sentiment Analysis, Natural Language Processing (NLP), Naive Bayes Algorithm, Support Vector Machine(SVM), Logistic Regression.
Scope of the Article:  Machine Learning