Multilingual projection techniques with DBpedia for knowledge enrichment지식 심화를 위한 디비피디아 다언어 투영 기법

Cited 0 time in webofscience Cited 0 time in scopus
  • Hit : 861
  • Download : 0
Semantic link networks have been formed among data issued from various knowledge sources distributed around the globe through Linked Data, creating an environment that allows data reuse as well as information and knowledge sharing. However, there are limitations to extent of progress in Linked Data varies across languages due to the scale of relations among data grows rapidly as the fields that participate in Linked Data become diverse and as the specified ontologies become complex. This dissertation discusses a method for enriching knowledge by projecting multilingual Linked Data from DBpedia, the database that plays a key role in implementing Linked Data. First, parts that have maximal reusability for many different languages are extracted from DBpedia ontologies created solely from current English resources. Usability of ontologies is improved while meeting the demands of various languages by selecting essential semantic data among the languages. Second, this dissertation extends its research domain to compare the performance between technology based on dataset interfaces in various languages and technology based on individual languages. The proposed entity summarization method is achieved by projecting data extracted from multilingual datasets into a single joint space which is superior in performance to the state-of-the-art (at the time of this writing) technique for entity summarization that utilizes individual language-based approach that also requires additional external resources. In conclusion, this study demonstrates that projection of existing information among data in various languages can efficiently reveal the diversity and associations that are likely to be overlooked in the existing information extraction approach based on individual languages. Since multilingual projection-based information extraction complements the shortcomings of the individual language-based analysis, the multilingual method is expected to improve other extensive applications through combinations with various natural language processing technologies.
Advisors
Choi, Key-Sunresearcher최기선researcher
Description
한국과학기술원 :전산학부,
Publisher
한국과학기술원
Issue Date
2016
Identifier
325007
Language
eng
Description

학위논문(박사) - 한국과학기술원 : 전산학부, 2016.8 ,[iv, 81 p. :]

Keywords

Semantic Web; Linked Data; Multilingual projection; Ontology; Entity summarization; 시맨틱 웹; 링크드 데이터; 다언어 투영; 온톨로지; 개체 요약

URI
http://hdl.handle.net/10203/222420
Link
http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=663201&flag=dissertation
Appears in Collection
CS-Theses_Ph.D.(박사논문)
Files in This Item
There are no files associated with this item.

qr_code

  • mendeley

    citeulike


rss_1.0 rss_2.0 atom_1.0