RoBERTa leads sentiment analysis, outperforming traditional models

Researchers explored various machine learning models for sentiment classification of movie reviews, comparing traditional methods with transformer-based approaches. The study utilized the IMDb dataset and evaluated models including Naive Bayes, Logistic Regression, SVM, LightGBM, LSTM, RoBERTa, and DistilBERT. RoBERTa achieved the highest accuracy at 93.02%, and a soft voting ensemble combining all models further enhanced performance, demonstrating the effectiveness of model ensembling for sentiment analysis. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Demonstrates superior performance of transformer models like RoBERTa over traditional methods for sentiment analysis, potentially guiding future NLP application development.

RANK_REASON Academic paper detailing comparative model performance on a specific task. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CL →

• DistilBERT
• IMDb dataset
• LightGBM
• Logistic Regression
• LSTM
• Naive Bayes
• RoBERTa

• paper
• product