Sentiment analysis for Product Reviews

Abstract

Sentiment analysis is an effective method for identifying text data and extracting the sentiment component. Every day, customers' reviews, opinions, suggestions, and tweets generate a high amount of unstructured data on shopping websites. Retailers can use aspect level analysis of this data to gain a better knowledge of their customers' expectations and then modify their policies accordingly. A innovative approach based on aspect level sentiment detection, which focuses on the item's features, is provided in this research. The work was implemented and validated on Amazon customer reviews (crawled data), where each review's aspect phrases were determined initially.