Papers by Bunyamin Sisman
CoRI: Collective Relation Integration with Data Augmentation for Open Information Extraction (2021.acl-long)
Copied to clipboard
| Challenge: | Existing methods to integrate extracted knowledge from the Web to knowledge graphs (KGs) however, the predictions are made independently, which can be mutually inconsistent. |
| Approach: | They propose a relation integration model that aligns free-text relations to relations in a target KG . they propose combining two stages to make independent predictions and a collective model that accesses all candidate predictions. |
| Outcome: | The proposed model outperforms baseline models on two datasets and improves AUC from .677 to .748 and from 1.716 to 1.780. |
LinkNBed: Multi-Graph Representation Learning with Entity Linkage (P18-1)
Copied to clipboard
| Challenge: | Knowledge graphs have emerged as an important model for studying complex multi-relational data. |
| Approach: | They propose a deep relational learning framework that learns entity and relationship representations across multiple graphs. |
| Outcome: | The proposed framework improves on the state-of-the-art relational learning approaches and identifies entity linkage across graphs. |