Sentiment Analysis of product reviews using Hadoop

Sentiment analysis or opinion mining is one amongst the most important tasks of information processing (Natural Language Processing). Sentiment analysis has gain abundant attention in recent years. There is a tendency to aim to tackle the matter of sentiment polarity categorization, that is one amongst the elemental issues of sentiment analysis. Therefore, a way that assigns scores indicating positive and negative opinion concerning the merchandise is planned. It uses Hadoop Distributed File system (HDFS) to store knowledge set and run on MapReduce design for performing sentiment analysis. A general method for sentiment polarity categorization is planned with elaborated method, descriptions knowledge utilized in this study square measure on-line product reviews collected from Amazon.com. Experiments for each sentence-level categorization and review-level categorization square measure performed with promising outcomes. Recommendation system provides the ability to grasp a person’s style and notice new, fascinating content. This would help the organization to decide where they lack, where they need to improve, based on the sentiment analysis carried out using Hadoop.