LANGUAGE CHARACTERISTIC EXTRACTION DEVICE, NAMED ENTITY EXTRACTION DEVICE, EXTRACTION METHOD, AND PROGRAM

摘要

In order to suitably accommodate differences among characteristics that should be considered on a language-specific basis, and enable implementation of a common named entity extraction as a processing system, a language characteristic extraction device 11 is equipped with a language characteristic extraction unit that: selects an abstraction rule from a set of abstraction rules which are common to a plurality of languages, said selection being made in accordance with a characteristic of a target language; defines a language-specific characteristic extraction rule for each of a plurality of target languages, said language-specific characteristic extraction rule being defined as an extraction method and an output condition for a specific feature conforming to the target language; references, with respect to a morphological analysis result for an input sentence, the language-specific characteristic extraction rule that has been defined with respect to the language of the input sentence and is for the purpose of extracting a feature pertaining to an expression or part of speech included in the morphological analysis resu extracts a feature in accordance with the language; and outputs the feature as a language characteristic extraction result.