Classical ML outperforms deep learning on IMDb sentiment analysis

A new research paper compares traditional machine learning techniques with deep learning models for sentiment classification using IMDb movie reviews. The study found that classical methods, specifically Support Vector Machines with TF-IDF features, achieved higher accuracy than deep learning models like BiLSTM. While deep learning models showed promise in capturing sequential data, classical approaches proved more effective given resource constraints and feature engineering. AI

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

IMPACT Demonstrates that classical machine learning can still be competitive for specific NLP tasks, especially with limited resources.

RANK_REASON The cluster contains an academic paper detailing a comparative analysis of machine learning approaches. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CL →

• IMDb
• Logistic Regression
• Naïve Bayes
• Support Vector Machine
• BiLSTM
• BiLSTM with Attention
• Martin Clinton Tosima Manullang

• paper
• other