tweeteval

AI-powered tweet sentiment classification dashboard with in-browser ML inference, built with Logistic Regression, TF-IDF, and Chart.js on GitHub Pages. Zero server required.

Three-class TweetEval sentiment classification with token and VADER features, model comparison, and class-level error analysis.

Research on uncertainty quantification for social media analysis: emotion tagging, sentiment classification, and crisis triage under explicit error budgets.

Twitter sentiment classification with BERT/RoBERTa: LoRA parameter-efficiency sweep and cross-model error analysis on TweetEval