Natural Language Processing for Automated Sentiment Analysis of Consumer Reviews

Nur Aisyah Rahman

Natural Language Processing for Automated Sentiment Analysis of Consumer Reviews

Author(s)	Dr. Nur Aisyah Rahman
Country	United States
Abstract	Consumer reviews contain valuable information about product quality, service performance, customer expectations, and post-purchase experiences. However, the volume, linguistic variability, and unstructured nature of online reviews make manual analysis increasingly impractical. Natural language processing provides computational methods for automatically identifying positive, negative, and neutral opinions and converting large collections of textual feedback into structured evidence for managerial decision-making. This study develops a reproducible simulation-based framework for automated sentiment analysis of consumer reviews. Four classification approaches—Multinomial Naive Bayes, linear support vector machine, bidirectional long short-term memory, and fine-tuned Bidirectional Encoder Representations from Transformers—are evaluated using 12,000 synthetic consumer reviews distributed across electronics, household products, online retail services, and personal-care products. Sentiment labels are generated through a controlled latent-score mechanism that incorporates positive and negative expressions, negation, intensifiers, contrastive conjunctions, aspect conflicts, and contextual modifiers. No private customer records, proprietary platform data, or real consumer identities are used. The simulated findings show that fine-tuned BERT achieves the strongest performance, with 92.6% accuracy and a macro F1 score of 92.4%. The bidirectional LSTM records a macro F1 score of 88.7%, followed by the linear support vector machine at 84.9% and Multinomial Naive Bayes at 78.6%. Error analysis indicates that mixed opinions, implicit sentiment, sarcasm, domain-specific expressions, and neutral–negative boundary cases remain difficult. The study concludes that transformer-based NLP can improve sentiment classification, but reliable organizational use requires transparent data governance, domain validation, class-sensitive evaluation, explainability, privacy protection, and continuing human review.
Keywords	natural language processing, sentiment analysis, consumer reviews, text classification, BERT, bidirectional LSTM, opinion mining, customer analytics.
Field	Engineering
Published In	Volume 1, Issue 2, March-April 2019
Published On	2019-03-01

View / Download PDF File

About AIJRI Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+1-872-302-3851	editor@aijri.com

American International Journal of Research and Innovation

A Widely Indexed Open Access Peer Reviewed Multidisciplinary Bi-monthly Scholarly International Journal

Natural Language Processing for Automated Sentiment Analysis of Consumer Reviews

Share this