We introduce a novel supervised learning-based approach to automate the mapping of web table columns to classes from an ontology. Our model is easy to learn and we only utilize the features found on the tables itself. Primarily we exploit the behavior of binary relations of columns in the tables to derive mappings. Secondarily we utilize column headers and column cell values. We outperform the current state of the art by the recall and F1 measure.
Details of Publications
Automating web table columns to knowledge base mapping using translation embedding, 2019 ICSC, Under review
Link to Paper - https://ieeexplore.ieee.org/abstract/document/9031447
Links to Resources
https://www.github.com/kavinduchamiran/fyp2
Team Members - Amila Rukshan
