Cross domain analyzer to acquire review proficiency in big data

Abstract Sentiment analysis is the pre-eminent technology for extracting relevant information in the data domain. In this paper, a cross-domain sentimental classification approach, the cross-domain analyzer (CDA), is proposed, which will extract positive words and replace their synonyms to escalate polarity. Additionally, the approach blends two different domains and detects all self-sufficient words. This is executed on Amazon datasets, in which two different domains are trained to analyze the sentiments of the reviews in the other domain. The proposed approach contributes promising results in the cross-domain analysis, and an accuracy of 92% is achieved. In BOMEST, the CDA improves precision and recall by 16% and 7%, respectively.