Software Research Laboratory

areas for Research & Development

  Machine Translation

• Natural language processing and understanding
• Machine translation for documents and dialogues
• Statistical machine translation
• Interactive interface for authoring tools
• Dialogue understanding and management
• Knowledge management for language resources

Machine Translation is a technology to translate a certain language to another language using a computer. These days, there is no physical limit between countries due to the development of various computer networking. However, there is one remaining limitation which is a "language". Currently, machine translation is being actively investigated in the goal of being used in a military purpose, a web globalization, a software as well as a document localization and dialog style translation such as a caption for broadcasting and chatting translation. The importance of Machine Translation technology is being highly emphasized since it is one of the novel technologies which will give a huge effect on human life in the near future.

  Speech Interface Technology

• Speech analysis and modeling
• LVCSR (Large Vocabulary Continuous Speech Recognition)
• Embedded large vocabulary speech recognition
• Conversational/Spontaneous speech recognition
• Noise robust speech recognition
• Microphone array processing
• HMM based speech synthesis

Speech Interface is the next generation interface which makes a computer understand what the person says and also to exchange information in the smooth conversation between a computer and a person. Recently the speech interface has become one of the essential elements in the IT industry such as intelligent robots, telematics, the next generation PC, and digital-home, which are the major targets of government’s promotions. Moreover, speech interface plays a role of core technology for mobile web information service and also can be applied to English education service for the conversation training. Speech interface is expected to be combined with many application areas such as national defense, medical services, etc. Henceforth Speech Interface technology will grow into multimodal interface with a united input device such as voice, pen, mouse, and gesture. A high level interactive Speech Interface will be soon attainable by incorporating circumstantial information and speaker's intention.

  Knowledge Mining Technology

• Mobile intelligent web search
• Mobile search with speech recognition
• Open web question answering technology
• Semantic Web search
• Crawling & analysis of web/mobile Contents
• Topic-based document classification
• Contextual online advertisement

Knowledge Mining is a research area for essential knowledge processing in the future ubiquitous computing environment and next generation web environment. It includes the state-of-the-art technologies such as natural language processing, text mining, question/answering as well as dialogue processing, knowledge consulting, multimedia semantic processing, semantic web, and ontology. The Knowledge Mining Research Team is promoting emphatically on the development of open web question/answering technology and mobile intelligent search technology with speech recognition as well as personal media management and context-based online/mobile advertisement.

  Automatic Speech Translation

• Spontaneous speech recognition
• Spoken language translation
• Dialog-style speech synthesis
• Spoken language understanding
• Multi-modal interface
• Speech and language database

Automatic Speech Translation is a technology to let a person have a conversation with people from different countries using his/her own native language. It is composed of three element techniques which are speech recognition which transforms voice to letters, machine translation which translates letters into another language, and voice synthesis which converts the letters to the voice. There is an extensive and diversified applied sphere such as a role of assistant for a trip abroad realtime interpretation for an international congress. Automatic Speech Translation will make a revolutionary upheaval variation to the way of living possible as the level of completion increases.