Reproduce “Large Language Models as Explainable AI Ensemble Aggregators for Business Review Sentiment Analysis: A Comparative Study with Classical Ensembles” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Konstantinos I. Roumeliotis, Dionisis Margaris, Dimitris Spiliotopoulos, Costas Vassilakis
The authors present LLAMA_AGG, which combines a review with star predictions and confidence scores from Logistic Regression, Support Vector Machine, Naïve Bayes, and BERT to produce a rating and rationale. On Yelp’s naturally distributed 10,000-review test set, it achieves macro-F1 of 0.6720, versus 0.6542 for Stacking. This lead is modest and does not establish faithful reasoning or generalization beyond Yelp.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.