Hiroshi Masuichi

The Parallel Grammar project

We report on the Parallel Grammar (ParGram) project which uses the XLE parser and grammar development platform for six languages: English, French, German, Japanese, Norwegian, and Urdu.

A bootstrapping method for extracting bilingual text pairs

This paper proposes a method for extracting bilingual text pairs from a comparable corpus. The basic idea of the method is to apply bootstrapping to an existing corpus-based cross-language information retrieval (CLIR) approach. We conducted preliminary tests with English and Japanese bilingual corpora. The bootstrapping method led to much better results for the task of extracting translation pairs compared with a corpus-based CLIR method without boot-strapping, and the extracted translation pairs could be useful training data for improving results of the corpus-based CLIR method.

An autonomous decentralized system for constraint-oriented problem solving involving continuous and fuzzy variables

A novel way of solving constraint satisfaction problems by using autonomous decentralized systems is introduced. First, the componential constraints are further decomposed into their structural and elastic parts. Second, the elastic parts are treated by interactive dynamics among autonomous decentralized problem solving processes. Third, the structural part is treated in a centralized logic based symbol processing subsystem which works concurrently and interactively with the autonomous subsystems for elastic parts. Fourth, this system architecture, is implemented on transputers by using Occam and is applied to scheduling and mechanical design problems showing the effectiveness of the approach.<<ETX>>

The Japanese lexical transducer based on stem-suffix style forms

A Lexical Transducer (LT) as defined by Karttunen, Kaplan, Zaenen 1992 is a specialized finite state transducer (FST) that relates citation forms of words and their morphological categories to inflected surface forms. Using LTs is advantageous because the same structure and algorithms can be used for morphological analysis (stemming) and generation. Morphological processing (analysis and generation) is computationally faster, and the data for the process can be compacted more tightly than with other methods. The standard way to construct an LT consists of three steps: (1) constructing a simple finite state source lexicon LA which defines all valid canonical citation forms of the language; (2) describing morphological alternations by means of two-level rules, compiling the rules to FSTs, and intersecting them to form a single rule transducer RT; and (3) composing LA and RT.