Identifying Bilingual Topics in Wikipedia for Efficient Parallel Corpus Extraction and Building Domain-Specific Glossaries for the Japanese-English Language Pair
This paper presents an approach and its implementation as a software toolset for examining what portion of the multilingual content of Wikipedia is viable for harvesting bilingual data in order to build parallel corpora and domain-specific glossaries. An algorithm is presented which analyzes the link topology of topics and subtopics and the co-occurance in another language. This algorithm is implemented in the Python language and can be used to examine an arbitrary number of topics for Japanese-English as well as other language pairs with minor adjustements. The goal of the toolchain is ease of use and transparency as well as flexibility towards language combinations. The findings of a test with several thousands topics is presented as a showcase. The toolchain is open source under the Creative Commons license.
@InProceedings{WLOKA18.3, author = {Bartholomäus Wloka}, title = {Identifying Bilingual Topics in Wikipedia for Efficient Parallel Corpus Extraction and Building Domain-Specific Glossaries for the Japanese-English Language Pair}, booktitle = {Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC 2018)}, year = {2018}, month = {may}, date = {7-12}, location = {Miyazaki, Japan}, editor = {Reinhard Rapp and Pierre Zweigenbaum and Serge Sharoff}, publisher = {European Language Resources Association (ELRA)}, address = {Paris, France}, isbn = {979-10-95546-07-8}, language = {english} }
Summary of the paper