Online Product Recommendation System Using Sentiment Analysis and Spam Filtering

Abstract

E-commerce has recently concentrated on the rise of internet retail platforms and can be recommended by multiple consumers who refer to product review opinions to pick their goods and products from their purchasing experience. But, if current product evaluation programs include customers' problems (eg. cost, power, structure, function, etc.) conventional recommendation systems do not recommend alternative products, it is also difficult to meet user requirements. In this paper, we suggest therefore a new product recommendation framework that analyses two forms of information: information on grievances and information on satisfaction from e-commerce review comments. Alternative things can be recommended to meet the needs and can resolve the product issue details as you browse. In this article, we would identify the extraction of complaint information by removing both negative information and positive information from product feedback and clarify alternative product recommendation approaches for complaints resolution and check the efficacy of complaint information extraction and alternative product recommendations.