Reproduce “ReviewGraph: A Knowledge Graph Embedding Based Framework for Review Rating Prediction with Sentiment Features” 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
A.J.W. de Vink, Natalia Amat-Lefort, Lifeng Han
ReviewGraph turns hotel reviews into an inspectable graph for rating prediction. It extracts subject–predicate–object triples, scores their sentiment, links topics to reviews and hotels, and combines graph embeddings with sentiment features in classifiers. In an 80/20 test, its best configuration reached 0.2829 Cohen’s kappa and 0.5404 accuracy, below Bag-of-Words Logistic Regression; the graph’s value is representation, not predictive superiority.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.