DEEP LEARNING BASED SENTIMENT ANALYSIS USING REVIEWS RESPONSE FOR E-SHOPPING WEBSITES



Authors

  • Santosh Gaikwad1, R. Shanmugan2, Bharti Gawali3, Chitra Desai4 , Suhas Mache5

DOI:

https://doi.org/10.15282/jmes.17.1.2023.10.0759


Keywords:

Sentimental Analysis, Deep Leaning, Product Review, Internet, Amazon, E-shopping, opinion, classification.


Abstract

Sentimental analysis is the most important task in Natural language Processing. The Growth of internet increases day by day. Internet has become very popular resource for information gathering. Normal human beings depend on internet for their day to day activity. The users express their opinion about the product, services of website and its limitation. Millions of users discussed their opinions as a valuable platform for tracking and analyzing opinion and sentiments. For purchasing the any items or products from online market, what other peoples posted review and their opinion about services play an important role. What people rating our product and their opinion regarding product and services is also part of research and development department of concern website. Sentimental analysis gives important evidence for decision making in various domains. The deep learning is the prominent technology in the current era. The main objectives of our research are to recognize the sentiments such as negative, positive and neutral regarding the e-shopping websites and products. This research aimed at a large dataset for four different e-shopping websites which allow the review based services. For the finding of the score of the each word from review, we are using the “Sentiwordnet dictionary”. From this research, It is observed that the pre-processing of the review based dataset is mostly affecting the quality of the sentimental analysis. The classification of sentimental analysis such as positive, negative and neutral has been done with the deep learning approach.



Published

2024-04-29

How to Cite