Unsupervised Text Annotations

Tanya Braun, Felix Kuhr, Ralf Möller

Abstract

We introduce the unsupervised text annotation model UTA, which iteratively populates a document-specific database containing the related symbolic content description. The model identifies the most related documents using the text of documents and the symbolic content description. UTA extends the database of one document with data from related documents without ignoring the precision.

Original languageEnglish
Title of host publicationProceedings of the 6th Workshop on Dynamics of Knowledge and Belief (DKB-2017) and the 5th Workshop KI Kognition (KIK-2017) co-located with 40th German Conference on Artificial Intelligence (KI 2017)
Number of pages8
Volume1928
PublisherCEUR-WS.org
Publication date01.09.2017
Pages23-30
Publication statusPublished - 01.09.2017
Event6th Workshop on Dynamics of Knowledge and Belief and the 5th Workshop KI and Kognition - Dortmund, Germany
Duration: 26.09.201726.09.2017
Conference number: 130603

DFG Research Classification Scheme

  • 409-06 Information Systems, Process and Knowledge Management

Cite this