RelationalCoder: Rethinking Complex Tables via Programmatic Relational Transformation
Semi-structured tables, with their varied layouts and formatting artifacts, remain a major obstacle for automated data processing and analytics.To address these challenges, we propose RELATIONALCODER, which uniformly converts semi-structured tables into relational data, enabling smooth integration with the rich ecosystem of data processing and analytics tools.By leveraging SQL code, RELATIONAL-CODER prevents schema errors and markedly improves normalization quality across multiple relational tables.
Authors
- Huailiang Peng (ORCID: https://orcid.org/0009-0000-4943-8853)
- Yanan Cao (ORCID: https://orcid.org/0000-0003-3534-1094)
- Yue Hu (ORCID: https://orcid.org/0000-0002-1116-7303)
- Haoyu Dong (ORCID: https://orcid.org/0009-0007-5003-6801)
Institutions
- Institute of Information Engineering (CN)
Publication Details
- Published
- 2025-01-01
- DOI
- https://doi.org/10.18653/v1/2025.acl-long.89
- Citations
- 1
- Primary Topic
- Advanced Database Systems and Queries
- Type
- article
- Field-Weighted Citation Impact
- 1.49
Funders
- National Natural Science Foundation of China