2026-10-11T14:26:56Z https://meral.edu.mm/oai

oai:meral.edu.mm:recid/4724 2022-03-24T23:12:53Z 1582963302567:1597824273898 user-ucsy

A Proposal on Statistical Language Model Building for Under-Resourced Languages Nakajima, Hideharu Corpora must be developed for realizing statisticallanguage models; however, it is labor-intensive workand is very expensive. Often corpora resources arescant for such a domain as new language, new task, andnew application. This situation is called “underresourced.”Myanmar must be a one of under-resourcedlanguages for Japanese computational linguists, andvice versa. For both languages, mother tongues are“well-resourced.” Fortunately, as Japanese andMyanmar show the same Subject-Object-Verb (SOV)word order, higher translation quality can be expectedbetween these two languages. Thus, Japanese andMyanmar can be mediator (or hub) languages to otherlanguages each other. This paper proposes thereutilization of well-resourced domain data to offsetunder-resourced domain data. As a feasible example,this paper introduces the essence of my past research [1,2] on model adaptation with machine translated text. 2012-02-28 http://hdl.handle.net/20.500.12678/0000004724 https://meral.edu.mm/records/4724