Sentiment Classification Using Document Embeddings Trained with Cosine Similarity

ACL 2019 Tan ThongtanTanasanee Phienthrakul

In document-level sentiment classification, each document must be mapped to a fixed length vector. Document embedding models map each document to a dense, low-dimensional vector in continuous vector space... (read more)

PDF Abstract

Evaluation Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK COMPARE
Sentiment Analysis IMDb NB-weighted-BON + dv-cosine Accuracy 97.4 # 1