A Cascade Dual-Decoder Model for Joint Entity and Relation Extraction
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2021
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866910455435362304 |
|---|---|
| author | Cheng, Jian Zhang, Tian Zhang, Shuang Ren, Huimin Yu, Guo Zhang, Xiliang Gao, Shangce Ma, Lianbo |
| author_facet | Cheng, Jian Zhang, Tian Zhang, Shuang Ren, Huimin Yu, Guo Zhang, Xiliang Gao, Shangce Ma, Lianbo |
| contents | In knowledge graph construction, a challenging issue is how to extract complex (e.g., overlapping) entities and relationships from a small amount of unstructured historical data. The traditional pipeline methods are to divide the extraction into two separate subtasks, which misses the potential interaction between the two subtasks and may lead to error propagation. In this work, we propose an effective cascade dual-decoder method to extract overlapping relational triples, which includes a text-specific relation decoder and a relation-corresponded entity decoder. Our approach is straightforward and it includes a text-specific relation decoder and a relation-corresponded entity decoder. The text-specific relation decoder detects relations from a sentence at the text level. That is, it does this according to the semantic information of the whole sentence. For each extracted relation, which is with trainable embedding, the relation-corresponded entity decoder detects the corresponding head and tail entities using a span-based tagging scheme. In this way, the overlapping triple problem can be tackled naturally. We conducted experiments on a real-world open-pit mine dataset and two public datasets to verify the method's generalizability. The experimental results demonstrate the effectiveness and competitiveness of our proposed method and achieve better F1 scores under strict evaluation metrics. Our implementation is available at https://github.com/prastunlp/DualDec. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2106_14163 |
| institution | arXiv |
| publishDate | 2021 |
| record_format | arxiv |
| spellingShingle | A Cascade Dual-Decoder Model for Joint Entity and Relation Extraction Cheng, Jian Zhang, Tian Zhang, Shuang Ren, Huimin Yu, Guo Zhang, Xiliang Gao, Shangce Ma, Lianbo Computation and Language Artificial Intelligence In knowledge graph construction, a challenging issue is how to extract complex (e.g., overlapping) entities and relationships from a small amount of unstructured historical data. The traditional pipeline methods are to divide the extraction into two separate subtasks, which misses the potential interaction between the two subtasks and may lead to error propagation. In this work, we propose an effective cascade dual-decoder method to extract overlapping relational triples, which includes a text-specific relation decoder and a relation-corresponded entity decoder. Our approach is straightforward and it includes a text-specific relation decoder and a relation-corresponded entity decoder. The text-specific relation decoder detects relations from a sentence at the text level. That is, it does this according to the semantic information of the whole sentence. For each extracted relation, which is with trainable embedding, the relation-corresponded entity decoder detects the corresponding head and tail entities using a span-based tagging scheme. In this way, the overlapping triple problem can be tackled naturally. We conducted experiments on a real-world open-pit mine dataset and two public datasets to verify the method's generalizability. The experimental results demonstrate the effectiveness and competitiveness of our proposed method and achieve better F1 scores under strict evaluation metrics. Our implementation is available at https://github.com/prastunlp/DualDec. |
| title | A Cascade Dual-Decoder Model for Joint Entity and Relation Extraction |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2106.14163 |