Not logged in.

Contribution Details

Type Conference or Workshop Paper
Scope Discipline-based scholarship
Published in Proceedings Yes
Title Aligning Knowledge Base and Document Embedding Models using Regularized Multi-Task Learning
Organization Unit
Authors
  • Matthias Baumgartner
  • Wen Zhang
  • Bibek Paudel
  • Daniele Dell' Aglio
  • Huajun Chen
  • Abraham Bernstein
Presentation Type paper
Item Subtype Original Work
Refereed No
Status Published in final form
Language
  • English
ISBN 978-3-030-00670-9
Page Range 21 - 37
Event Title ISWC 2018
Event Type conference
Event Location Monterey
Event Start Date October 8 - 2018
Event End Date October 12 - 2018
Series Name International Semantic Web Conference
Number 11136
Place of Publication Cham
Publisher Springer
Abstract Text Knowledge Bases (KBs) and textual documents contain rich and complementary information about real-world objects, as well as relations among them. While text documents describe entities in freeform, KBs organizes such information in a structured way. This makes these two information representation forms hard to compare and integrate, limiting the possibility to use them jointly to improve predictive and analytical tasks. In this article, we study this problem, and we propose KADE, a solution based on a regularized multi-task learning of KB and document embeddings. KADE can potentially incorporate any KB and document embedding learning method. Our experiments on multiple datasets and methods show that KADE effectively aligns document and entities embeddings, while maintaining the characteristics of the embedding models.
Digital Object Identifier 10.1007/978-3-030-00671-6_2
Other Identification Number merlin-id:16389
PDF File Download from ZORA
Export BibTeX
EP3 XML (ZORA)
Additional Information ISBN: 978-3-030-00671-6 (E)