Deep sentiments extraction for consumer products using NLP-based technique

Abstract

The growth in the field of e-commerce and product availability over the Internet is the higher availability of the consumable items is making the customers seek for higher quality and comparative price points. The primary reason for this ambiguity is the lack of in hand experience for the customers before the purchase. The customers mostly tend to rely on the feedbacks of the other buyers. The feedbacks on the products are often made in thousands in numbers, and it is difficult for the potential buyers to decide by looking into these feedbacks or reviews. Thus the demand of the modern research is to automate the process for extracting the true feedback matching their needs based on usage or price or location constraints. The feedback or the review system can be easily manipulated by the incorrect feedbacks. Hence it is important to reduce the influence of those feedbacks during extracting the overall sentiment of any product. Also, yet another challenge is that most of the feedbacks are not in formal English, thus making it difficult to extract the accurate feedback. This work proposes a novel-automated frame for extracting the deep sentiments from the reviews or the feedbacks on e-commerce websites. Another major outcome of this work is to detect the false reviews and making the sentiment true for any decision making. The research work generates a trustable sentiment extraction process to justify the need of true feedbacks for customer decision making.