Papers by Takuya Hiraoka
Cross-Sentence N-ary Relation Extraction using Lower-Arity Universal Schemas (D19-1)
Copied to clipboard
| Challenge: | Existing approaches to extract n-ary relations from text are limited to binary relations. |
| Approach: | They propose to learn relation representations of lower-arity facts from decomposing higher-arities . they conduct experiments with datasets for ternary relation extraction . |
| Outcome: | The proposed method improves the performance of n-ary relation extraction methods compared to previous methods. |