Abstract
This paper introduces a novel late interaction mechanism for knowledge base question answering (KBQA) systems, combining Graphormer and transformer representations. We conducted extensive experiments, comparing various pooling mechanisms and configurations. Our results demonstrate significant improvements in F1-score compared to traditional baselines. Specifically, we found that attention pooling, in conjunction with linearized graph and question features alongside sub-graph representations, yields the best performance. Our study highlights the importance of advanced interaction mechanisms and the integration of diverse modalities in KBQA systems.- Anthology ID:
- 2024.textgraphs-1.12
- Volume:
- Proceedings of TextGraphs-17: Graph-based Methods for Natural Language Processing
- Month:
- August
- Year:
- 2024
- Address:
- Bangkok, Thailand
- Editors:
- Dmitry Ustalov, Yanjun Gao, Alexander Panchenko, Elena Tutubalina, Irina Nikishina, Arti Ramesh, Andrey Sakhovskiy, Ricardo Usbeck, Gerald Penn, Marco Valentino
- Venues:
- TextGraphs | WS
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 137–141
- Language:
- URL:
- https://aclanthology.org/2024.textgraphs-1.12
- DOI:
- Cite (ACL):
- Mayank Rakesh, Parikshit Saikia, and Saket Shrivastava. 2024. TIGFORMER at TextGraphs-17 Shared Task: A Late Interaction Method for text and Graph Representations in KBQA Classification Task. In Proceedings of TextGraphs-17: Graph-based Methods for Natural Language Processing, pages 137–141, Bangkok, Thailand. Association for Computational Linguistics.
- Cite (Informal):
- TIGFORMER at TextGraphs-17 Shared Task: A Late Interaction Method for text and Graph Representations in KBQA Classification Task (Rakesh et al., TextGraphs-WS 2024)
- PDF:
- https://preview.aclanthology.org/autopr/2024.textgraphs-1.12.pdf