Language Grid : Building A Bridge for Language Barriers

On Tuesday, 4^th March, 2008, from 1 o’clock upto 3 o’clock in the afternoon, a discussion meeting between STEI team and Language Grid Project team from Japan was held. STEI team was represented by Dr. Arry Akhmad Arman (STEI lecturer and researcher from Information System and Human Machine Interface Research Group) and Dr. Ayu Purwarianti (STEI Computational Linguistics researcher from Graphic and Artificial Intelligence Research Group), while Language Grid team was represented by Prof. Toru Ishida (Project Leader, Professor of Department of Social Informatics in Kyoto University), Yohei Murakami PhD. (Researcher), Yoko Kubota (Laboratory Coordinator) and Arif Bramantoro PhD. (Student in Kyoto University).

At the beginning, Dr. Arry gave presentation about ‘Introducing R&D Activities in Language Technology in ITB’, to describe about language-related researches performed in ITB. Several cooperation with Japan, Belgium and German related to his Text-to-Speech researches was described, as well as some development of rule-based IndonesianEnglish translation software. However, Dr. Arry admitted that only few research activities are conducted together between Indonesian linguists and engineers. Dr. Ayu gave a quick and detailed explanation about her research work during her PhD. study in Toyohashi University of Technology in ‘Developing Cross Language System for Language Pairs with Limited Resources’. Her research would enable users to use pivot (intermediate) language for multi-source-target-languages translation scheme. In her work, she uses English as the pivot language in Indonesian to Japanese translation, for information retrieval and question&answer purposes. Later on, Prof. Ishida presented the Language Grid. The project team, consists of several research groups including NICT (National Institute of Information and Communication Technology), Universities, and NTT Corp., started developing a language infrastructure on the Internet called the Language Grid. This project is based on collaboration between industry, government, universities, and citizens. Practical systems will be released completely within three years from its starting time, April 2006.

Language Grid comprises of 4 service layers : Intercultural Collaboration Tools, Language Services, Language Resources and P2P Grid Infrastructure. Language Resources layer comprises of several main components : (1) Machine Translators, available for 8 language pairs : (Japanese, Chinese) (Japanese, Korean) (Japanese, English) (English, German) (English, Spanish) (English, French) (English, Italian) (English, Portuguese); (2) Morphological Analyzers, available in 11 languages : Japanese, Chinese, Korean, English, German, Spanish, French, Italian, Dutch, Russian, Bulgarian; (3) Bilingual dictionaries specific to certain fields and languages, available for Medical/Life science terms (Japanese, English), Disaster management terms (Japanese, Chinese, Korean, English, French, Spanish), Academic terms (Japanese, English); and the last one is (4) Language Service Ontology for wrapping the above language resources. The presentation was continued by Yohei Murakami PhD., describing Language Grid in a more technical nuance. The translation complexity for JapaneseEnglishGermany was shown, thereby it is suggested to develop direct translation process for JapaneseIndonesian, to reduce the performance and accuracy degradation. Further details can be read by visiting NICT Language Grid Project Web site (http://langrid.nict.go.jp/), Language Grid Operation Center Web site (http://langrid.org/operation/) run by Kyoto University, and Language Grid Association Web site (http://langrid.org/association/indexe.html).

The presentations went on till it came to the end: then continued with discussion about any possible partnerships. According to Prof. Ishida and Dr. Arry, possible partnerships would be some joint-researches on building new/modification/combination of Components for Language Grid service layers; providing IndonesiaJapanese machine translation, morphological analyzer, and bilingual dictionaries with some more terms; and also by adding more end-users to the Language Grid community.