The data set provides two emotion lexicons that manually labeled by three well-instructed coders to calculate the emotion polarity of JD reviews in Chinese. Specifically, the negative lexicon (jd_negative_terms.txt) contains 1, 589 unique negative terms and the positive lexicon (jd_positive_terms.txt) contains 1135 unique positive terms.
Please cite the following paper:
Di Weng and Jichang Zhao. Positive emotions help rank negative reviews in e-commerce. arXiv:2005.09837, 2020.
Any issues please feel free to contact jichang@buaa.edu.cn for more information.