When Hearst Is not Enough: Improving Hypernymy Detection from Corpus with Distributional Models
This work addresses hypernymy detection for natural language processing, offering an incremental improvement by handling specific sparsity cases.
The paper tackles hypernymy detection by quantifying cases where pattern-based methods fail due to sparsity and proposes a complementary framework combining pattern-based and distributional models, achieving competitive improvements on benchmark datasets.
We address hypernymy detection, i.e., whether an is-a relationship exists between words (x, y), with the help of large textual corpora. Most conventional approaches to this task have been categorized to be either pattern-based or distributional. Recent studies suggest that pattern-based ones are superior, if large-scale Hearst pairs are extracted and fed, with the sparsity of unseen (x, y) pairs relieved. However, they become invalid in some specific sparsity cases, where x or y is not involved in any pattern. For the first time, this paper quantifies the non-negligible existence of those specific cases. We also demonstrate that distributional methods are ideal to make up for pattern-based ones in such cases. We devise a complementary framework, under which a pattern-based and a distributional model collaborate seamlessly in cases which they each prefer. On several benchmark datasets, our framework achieves competitive improvements and the case study shows its better interpretability.