Integrating User Feedback under Identity Uncertainty in Knowledge Base Construction

Ari KobrenNicholas MonathAndrew McCallum.



This paper develops a framework for integrating user feedback under identity uncertainty in knowledge bases.
Users have tremendous potential to aid in the construction and maintenance of knowledges bases (KBs) through the contribution of feedback that identifies incorrect and missing entity attributes and relations. However, as new data is added to the KB, the KB entities, which are constructed by running entity resolution (ER), can change, rendering the intended targets of user feedback unknown–a problem we term identity uncertainty. In this work, we present a framework for integrating user feedback into KBs in the presence of identity uncertainty. Our approach is based on having user feedback participate alongside mentions in ER. We propose a specific representation of user feedback as feedback mentions and introduce a new online algorithm for integrating these mentions into an existing KB. In experiments, we demonstrate that our proposed approach outperforms the baselines in 70% of experimental conditions.


title={Integrating User Feedback under Identity Uncertainty in Knowledge Base Construction},
author={Ari Kobren and Nicholas Monath and Andrew McCallum},
booktitle={Automated Knowledge Base Construction (AKBC)},
Gold Sponsors
Silver Sponsors
Bronze Sponsors
Chan Zuckerberg Initiative Facebook Google
Diffbot Oracle Corporation NEC
Elsevier Kenome