Amazon Reviews for Computer Accessories: A Curated Dataset for Sentiment Analysis

Dataset Name: [Customer Product Review] Amazon Reviews for Computer Accessories: A Curated Dataset for Sentiment Analysis (2021-Present)Description:This dataset contains a curated collection of user reviews for computer accessories scraped from Amazon. It was specifically assembled for the development and evaluation of machine learning and deep learning models for sentiment analysis tasks. The data provides insights into contemporary customer opinions, preferences, and complaints within the technology product sector.The collection process utilized web scraping techniques with Beautiful Soup to gather reviews posted from 2021 onward, ensuring the data reflects recent market trends and user language. After an extensive data cleaning and deduplication process, the final dataset comprises 29,755 unique review entries.Each entry includes key metadata such as the review text, star rating, reviewer information, and other relevant details, making it a valuable resource for natural language processing (NLP) research.This dataset was used in a comparative study that implemented and evaluated a range of sentiment classification models, including:• Deep Learning Models: Convolutional Neural Networks (CNNs) and Bidirectional Encoder Representations from Transformers (BERT).• Traditional Machine Learning Models: Support Vector Machines (SVM) and Logistic Regression.• Explainable AI (XAI): The study also employed LIME (Local Interpretable Model-agnostic Explanations) to interpret model predictions and enhance transparency.The results demonstrated the superior performance of the BERT model, achieving an accuracy of 86.01% and a balanced F1-score of 88%, establishing it as a robust choice for understanding sentiment in customer feedback.Potential Research Applications:• Sentiment Analysis and Opinion Mining• Comparative Analysis of NLP Models (Traditional ML vs. Deep Learning)• Explainable AI (XAI) in NLP• Customer Behavior and Market Trend Analysis• Benchmarking for New Classification Algorithms Author: MD Abdullah Al Symum; Rafiatul Zannah; Subarna Yeasmin Sheemu;Affiliation: Department of Computer Science and Engineering, Brac University, Dhaka, Bangladesh;Correspondence: [e-mail]; [e-mail];[e-mail]; Paper DOI: https://doi.org/10.1109/ICCIT60459.2023.10441344 Paper Link: https://ieeexplore.ieee.org/document/11022600