We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
SEMI-AUTOMATIC ONTOLOGICAL ALIGNMENT OF DIGITIZED BOOKS PARALLEL CORPORA.
- Authors
LAUKAITIS, Algirdas; LAUKAITYTĖ, Neda
- Abstract
In this paper, we present a method for general ontology management integration with an alignment of digitized books paraphrase corpus, which have been compiled from bilingual parallel corpus. We show that our method can improve ontology development and consistency checking when we add semantic parsing and machine translation to the process of general knowledge management. Additionally, we argue that the focus on one’s favorite books gives a factor of gamification for knowledge management process. A new formalism of semantic parsing ontological alignments is introduced and its use for ontology development and consistency checking is discussed. It is shown that existing general ontologies requires much more axioms than it is currently available in order to explain unaligned content of books. Proactive learning approach is suggested as part of the solution to improve development of ontology predicates and axioms. WordNet, FrameNet and SUMO ontologies are used as a starting knowledge base of paraphrase corpus semantic alignment method
- Subjects
CORPORA; MACHINE translating; NATURAL language processing; KNOWLEDGE management
- Publication
Science: Future of Lithuania / Mokslas: Lietuvos Ateitis, 2021, Vol 13, p1
- ISSN
2029-2341
- Publication type
Article
- DOI
10.3846/mla.2021.15034