Current Research Project
Dialog System
  • Spoken Language Understanding (SLU)


  • Spoken Dialog Management (SDM)


  • Multimodal Dialog Management (MDM)


  • Statistical Dialog Simulation (DS)


  • Dialog System Workbench (DialogStudio)


    		•
    Search and Semantic web
  • Semi-supervised IE (IE)


  • Associative Text Classification (BCAR)


  • Speech and Multimedia IR (SIR)


  • Ontology-based Content Recommendation (OCR)


    		•
    Computer Assisted Language Learning
  • Dialogue-based Computer Assisted Language Learning (DB-CALL)


  • Easy-Write (editing)


  • Comprehension Assistancei (editing)


  • Language Learning Game (editing)


    		•
    High Quality TTS
      We are developing a high-quality Text-To-Speech (TTS) system using sophisticated Natural Language Processing (NLP) techniques. In this research we combined NLP and concatenated speech synthesis technology for generating natural prosody in TTS. It includes language model, prosodic model and synthetic model. It presents a hybrid model for high-quality TTS system which integrates both statistical and rule-based method.
    Statistical Machine Translation
      We have been developing the statistical machine translation system for speech to speech translation. We focus our research on text-to-text translation task now, but we will include speech-to-speech translation among our research topics soon. We have an interest in building a translation model, decoding a word graph and combining statistical machine translation system and speech recognizer.


    Previous Research Project
    POMLIP
      POMLIP is an endeavor to develop Korean/English/Japanese/Chinese language processing engine which can be used in both speech and NLP applications.
    Natural Language QA
      We are developing a information retrieval and question answering system supporting multiple languages such as Korean, English, Japanese and Chinese. It aims to satisfy the information needs of desktop and mobile users by interpreting context information of users and deep analyzing text such as World-wide Web documents. It includes various components such as web robot, filter / classifier, information retrieval, cluster / summarizer, and question answering to satisfy user’s various needs.
    POSBIOTM
      We have been developing text mining systems for biological literature. The POSBIOTM/NER system aims to automatically identify different sets of biological entities in the text. The POSBIOTM/Event Extraction system aims to extracts various interactions among these biological entities. At present we are focusing on the POSBIOTM/W workbench. The workbench comprises a managing tool, a Named-entity Recognition Tool, an Event Extraction Tool and an annotation tool. The ultimate goal is to construct a workbench which provides appropriate tools in support of collecting, managing, creating, annotating and exploiting rich biomedical text resources.

     

    San 31, Hyoja-Dong, Pohang, 790-784, Korea
    Phone: +82 54 279 5581 | Fax: +82 54 279 2299