A Feature Based Opinion Mining for Product Reviews using Naive Bayes and K-Nearest Neighbor Classifiers
Minu P Abraham1, Udaya Kumar Reddy K R2

1Minu P Abraham, Assistant Professor, Department of CSE, NMAM Institute of Technology, Nitte, Deralakatte (Karnataka), India.
2Dr. Udaya Kumar Reddy K.R, Diploma in Computer Science and Engineering from Siddaganga Polytechnic, Tumkur, Bangalore University, Bangalore (Karnataka), India.

Manuscript received on 18 June 2019 | Revised Manuscript received on 25 June 2019 | Manuscript published on 30 June 2019 | PP: 2795-2801 | Volume-8 Issue-5, June 2019 | Retrieval Number: E7825068519/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© 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 explosive growth in the technology of web like social media, directs more number of people to express their sentiments, feedbacks or opinions towards the service, events, individuals, topics, products or social issues. Since the data get bigger extensively in everyday life, this makes so difficult for customers, manufactures as well as for end users of online websites to derive a conclusion on the accessibility of big data available as reviews. Opinion mining is referred to as a natural language processing technique that gives out the knowledge extraction of viewpoints or attitudes from the review texts. Feature-Based Opinion mining system aims to find the main aspects or features of a specified entity and the sentiment expressed on that entity by using natural language processing and AI techniques. Most of the studies have been conducted on aspect based opinion mining but not any of the particular works have justified to be adequate for assessing the critical or majorfactors. The major factors with respect to aspect based opinion mining are implicit or implied aspects, explicit or direct aspects and multiple aspect based. The Aspect based opinion mining by considering all these critical factors helps in analyzing the aspect of a particular entity and its sentiment more accurately. In this paper, an explicit aspect based opinion mining is carried out using naive bayes and K-nearest neighborclassifiers to generate more accurate opinions for product reviews. The end user can easily get opinions based on the particular aspect of the entity and the proposed work proves that the sentiment classification using naive bayes classifier provide with more accuracy than using K-nearest neighbor classifier.
Keywords: Aspect Extraction, Explicit Aspects, Online reviews, Sentiment Analysis, Text Mining

Scope of the Article: Text Mining