Polarity shifting has been a challenge to automatic sentiment classification. In this paper, we create a corpus which consists of polarity-shifted sentences in the product reviews, where both the sentimental words and shifting trigger words are annotated. In particular, we group the polarity-shifted sentence structures into five main categories, i.e., negation, contrastive transition, modality, implication, and irrelevance Evaluation shows the statistics on the agreement of the annotation and the distribution of the five categories of polarity shifting is given.
|Name||Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|
|Conference||14th Workshop on Chinese Lexical Semantics, CLSW 2013|
|Period||10/05/13 → 12/05/13|
- Theoretical Computer Science
- Computer Science(all)