Title: Machine learning-based sentiment analysis of Gujarati reviews

Authors: Parita Shah; Priya Swaminarayan

Addresses: Faculty of Engineering and Technology, Parul University, Vadodara, Gujarat, India; Computer Engineering Department, Gandhinagar Institute of Technology, Gujarat, India ' Faculty of IT and CS, Parul University, Vadodara, Gujarat, India; PIET-MCA, Parul University, Vadodara, Gujarat, India; BCA-Parul Institute of Computer Application, Parul University, Vadodara, Gujarat, India

Abstract: Opinion examination is the investigation of applied information in an articulation, like appraisals, assessments, sentiments, or perspectives toward a point, individual, or component. Positive, negative, and unbiased articulations are altogether conceivable. The authors of this exploration have built a dataset of Gujarati film audits and give the discoveries produced by the proposed calculation message in the wake of performing sentiment examination utilising a five different machine classifier. The authors fostered various datasets to test our calculation's capacities with different machine classifiers. This paper clarifies how information was gathered to shape a dataset, as well as Gujarati text pre-handling, include determination, and order techniques. According to the results of the investigation, all of the classifiers are performing brilliantly, generating overall precision greater than 75%, however KNN is unable to produce preferred precision above the others.

Keywords: N-gram; feature selection; sentiment evaluation; Gujarati language; film analysis; machine classifier.

DOI: 10.1504/IJDATS.2022.124763

International Journal of Data Analysis Techniques and Strategies, 2022 Vol.14 No.2, pp.105 - 121

Received: 29 Sep 2021
Accepted: 07 Feb 2022
Published online: 08 Aug 2022 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article