Apparatus for providing information using voice dialogue interface and method thereof

• Main Authors: LEE, JAE WON, LEE, HYE JEONG, SEO, HUI GYEONG, LEE, YEONG BEOM, PARK, CHAN MIN, KIM, JEONG SU
• Format: Patent
• Language: English; Korean
• Published: 08.10.2003
• Edition: 7
• Subjects: ACOUSTICS; CALCULATING; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; MUSICAL INSTRUMENTS; PHYSICS; SPEECH ANALYSIS OR SYNTHESIS; SPEECH OR AUDIO CODING OR DECODING; SPEECH OR VOICE PROCESSING; SPEECH RECOGNITION

PURPOSE: A device and a method for providing information by voice conversation interface are provided to search information desired by a user in real time and supply various and friendly conversation via a voice interface according to the taste of the user, by avoiding the repetition of same conversation through the conversation history of the user. CONSTITUTION: A device for providing information by voice conversation interface includes a voice recognition part(110) for receiving voice signals to recognize sound and converting the voice signals to sentences, a conversation management part(130), an information extracting part(140), a user modelling part(150), and a sentence generating part(160). The conversation management part is formed of an intension analyzing part(131) and an intension generating part(133), and carries out the analyzation on the intension of the voice signals of a user and the generation of the intension for the response to the user. The information extracting part searches online/offline information for any query information input of the user. The user modelling part analyzes the propensity of the user by the input information on the user conversation. The sentence generating part carries out the conversion from the intension information of the user to sentences. A sentence synthesizing part(171) synthesizes and outputs voice from the input sentences. 본 발명은 음성대화 인터페이스를 이용한 정보제공장치 및 그 방법에 관한 것으로, 음성신호를 문장으로 변환하는 음성인식부; 문장을 의미구조로 분석하는 문장분석부; 분석된 의미구조로부터 화행 또는 의도정보를 추출하고, 추출된 화행 또는 의도정보로부터 응답을 위한 화행 또는 의도를 생성하는 대화관리부; 화행과 의도정보로부터 응답을 위한 문장을 생성하는 문장생성부; 생성한 문장을 음성으로 합성하는 음성합성부; 응답에 필요한 정보를 인터넷으로부터 실시간으로 추출하는 정보추출부; 사용자의 성향을 분석하고 분류하는 사용자모델링부를 포함하므로, 사용자가 원하는 정보를 실시간으로 검색하여 사용자의 특성에 맞추어 다양한 대화를 음성인터페이스로 제공할 수 있다.