Abstract:
PURPOSE: An apparatus and method for storing/restoring SGML/XML entity is provided to prevent a waste of an auxiliary memory and to store SGML/XML document more rapidly and to restore entities stored in a database in various forms into the original entities in accordance with the form of each entity by sharing an entity declaration being referred to a plurality of documents. CONSTITUTION: A user interface(100) comprises a document input interface(110) requesting SGML/XML DTD and an input of a document to a SGML/XML document managing system(200), and a document output interface(120) requests a restore of a document by transmitting a storing portion identifier to the SGML/XML document managing system(200) and presents data of the document to a user. The elements of the SGML/XML document managing system(200) are described as follows. A SGML/XML parser(210) provides a verification of an error and a parsing result to a main memory as a tree form by performing a parsing with respect to the SGML/XML DTD and the document. A SGML/XML document storing device(220) receives the parsing result and stores an entity referring portion included in the document. An entity storing device(230) stores an entity declaration of the SGML/XML DTD. A SGML/XML entity managing device(240) receives the storing portion identifier from a client and returns data of an entity object corresponded to the storing portion identifier. A database system interface(250) is provided to interwork to a database system(300). The database system(300) charges a role of the lower storing system. A SGML/XML database(400) stores the document and the entity.
Abstract:
본 발명은 고립어 엔베스트 인식결과를 위한 발화검증 방법 및 장치에 있어서, 엔베스트 음성인식을 통해 단어수준에서 인식된 결과에 대해 엔베스트 발화검증을 통해서 신뢰도를 측정하고, 동적 시간 와핑에 의한 음소간 유사도를 측정하여 신뢰도와 유사도를 기반으로 음성인식의 수락/거절 또는 판단불가 등을 표시함으로써 보다 신뢰성 높은 음성인식이 가능하도록 한다. 엔베스트, 발화검증, 음성, 인식, 신뢰도, 유사도
Abstract:
PURPOSE: An errors correction apparatus in a successive word and voice recognition system and method thereof are provided to enable a user to directly modify the words in a word list by determining the ranking in the words. CONSTITUTION: A trie dictionary construction unit(110) constructs pronunciation rows for words of a voice recognition pronunciation dictionary as a trie dictionary. A similar pronunciation words searching unit(120) searches pronunciation similar with error words from a word list. The similar pronunciation word search unit outputs searched candidate words. An errors correction apparatus includes a voice recognition system.
Abstract:
PURPOSE: A sentential form apparatus of a voice recognition sentence and a method thereof are provided to identify a sentence structure in an automatic translation process and add the sentence structure. CONSTITUTION: A keyword set setting unit(110) sets up a keyword sest based on a learning sentence. A sentence structure identifying unit(108) identifies a sentence of a specific language through the keyword set. A punctuation marks addition unit(112) adds the punctuation mark to the identified sentence. An automatic translation unit(150) outputs voice of a preset language through automatic translation.
Abstract:
본 발명은 연속으로 발성되는 한국어 숫자음성에 대한 음성인식을 수행하여 그 인식 결과에 대해 혼동행렬과 신뢰도 치에 기반하여 다수의 음성인식 후보를 생성하는 기술에 관한 것으로, 혼동행렬은 숫자음성에 대해 오인식이 발생하는 숫자들로 구성되므로, 사전에 실험용 데이터베이스를 사용하여 인식을 수행한다. 또한 음성인식의 결과로 도출되는 숫자별 인식 점수인 통계적 우도를 그 단어의 지속시간 지수인 프레임 수로 나눈 로그 우도비를 신뢰도 치로써 사용하는 것을 특징으로 한다. 본 발명에 의하면, 음성인식 알고리즘에서 성능의 저하 없이 N-best를 생성하기 위해 사용하는 기억장치 사용량과 탐색시간을 절약함으로써 음성인식 엔진의 효율을 높일 수 있다. 음성인식, 숫자음성, N-best, 신뢰도치, 혼동행렬
Abstract:
PURPOSE: Interactive contents providing device and method in an e-book system are provided to interactive contents by editing the contents provided from CP. CONSTITUTION: CP(Contents Provider)(102,103) generates and provides the contents. A contents making device(110) creates interactive contents by manufacturing and editing the contents received from the CP. The contents making device provides to the interactive contents to the CP. Terminals(105-108,119) receive the interactive contents from the CP and provide the interactive contents to users. The interactive contents comprise script, object data and scene data.