Abstract

This dataset is a large-scale Chinese hotel review data set collected by Tan Songbo. The corpus size is 10,000 reviews. The corpus is automatically collected and organized from Trip.com.

Instructions:

This dataset is organized into 4 subsets, including: ChnSentiCorp-Htl-ba-2000 (balanced corpus, 1000 articles for positive and negative categories), ChnSentiCorp-Htl-ba-4000 (balanced corpus, 2000 articles for positive and negative categories), ChnSentiCorp-Htl-ba-6000 (balanced corpus, 3000 positive and negative categories) and ChnSentiCorp-Htl-unba-10000 (unbalanced corpus, 7000 positive categories).