Machine Learning Sentiment Analysis of Product Reviews based on Deep Embedding

Authors

  • K. M. Rayudu, K. Narendra, T. Sivaratna Sai, D. Karunamma, K. Venkata Ratnam

DOI:

https://doi.org/10.17762/msea.v71i4.1694

Abstract

Customer reviews are essential for guiding potential buyers toward making the best decisions. The direction of the evaluation clause is one of the main issues with several web usage retrieval strategies that were lastly suggested (e.g. normal or deviant). Deep learning has been proven to be a successful strategy for handling interpersonal problems. Unconsciously, a neural network finds a useful image without human assistance. Even though, the availability of vast information has a significant impact on the effectiveness of profound learning. We suggest a brand-new deep learning technique for categorising opinions of product evaluation with the most popular scores as insufficient tracking markers. The plan consists of two steps: analysing phrases' top-level ratings and adding a scoring layer. employing clearly identified words for controlled tuning at a level just above integrated. Long-term memory and convergent extractors are two recent network technique types that are examined. To validate the proposed methodology, we would create a repository containing 22,682 Amazon-labeled feedback statements and 2.2 million badly tagged feedback phrases. Observational data demonstrate the proposed scheme's effectiveness and superiority over measurements.

Downloads

Published

2022-12-31

How to Cite

K. M. Rayudu, K. Narendra, T. Sivaratna Sai, D. Karunamma, K. Venkata Ratnam. (2022). Machine Learning Sentiment Analysis of Product Reviews based on Deep Embedding. Mathematical Statistician and Engineering Applications, 71(4), 9201–9211. https://doi.org/10.17762/msea.v71i4.1694

Issue

Section

Articles