Active Learning for Entity Alignment
MCML Authors
Max Berrendorf
Dr.
* Former Member
Evgeny Faerman
Dr.
* Former Member
Volker Tresp
Prof. Dr.
Principal Investigator
Abstract
Max Berrendorf
Dr.
* Former Member
Evgeny Faerman
Dr.
* Former Member
Volker Tresp
Prof. Dr.
Principal Investigator
Abstract
In this work, we propose a novel framework for the labeling of entity alignments in knowledge graph datasets. Different strategies to select informative instances for the human labeler build the core of our framework. We illustrate how the labeling of entity alignments is different from assigning class labels to single instances and how these differences affect the labeling efficiency. Based on these considerations we propose and evaluate different active and passive learning strategies. One of our main findings is that passive learning approaches, which can be efficiently precomputed and deployed more easily, achieve performance comparable to the active learning strategies.
inproceedings
DL4G @WWW 2020
5th International Workshop on Deep Learning for Graphs at the International World Wide Web Conference. Taipeh, Taiwan, Apr 21, 2020.Authors
M. Berrendorf • E. Faerman • V. TrespLinks
Research Area
BibTeXKey: BFT20b