The Institute of Scientific and Industrial Research, Osaka University

contact
home
japanese

Department of Knowledge Science

Staff

  • Photo
    Prof.
    K. KOMATANI
  • Photo
    Assoc. Prof.
    K. KOZAKI
  • Photo
    Assis. Prof.
    R. TAKEDA
  • Photo
    Assis. Prof.
    K. HAYASHI

Content of research

Intelligence of machines, e.g., to talk with humans, is still under development. To realize human-friendly and helpful robots, the spoken dialogue function is indispensable. We study basic technologies on spoken dialogue systems, such as acoustic signal processing, natural language processing, machine learning, knowledge representation and so on, across several layers from acoustic signal processing to social interaction. We have also been involved in the ontology engineering, which organizes human knowledge and describes it in a machine-readable format.

Current Research Topics

  1. 1. Active-listening dialogue systems adaptive to situations and users
  2. 2. Inferernce and understanding based on generalization of knowledge graphs
  3. 3. Domain knowledge acquisition through dialogues
  4. 4. Ontology engineering and Linked Open Data (LOD)
  5. 5. Sound source localization based on deep neural networks
  6. 6. Natural language processing using embeddings of large-scale knowledge graphs

Figure / Graph

  • Fig.1
    Spoken dialogue robots interacting with multiple participants

VIDEO INTRODUCTION

The Institute of Scientific and Industrial Research, Osaka University

contact home japanese

Department of Knowledge Science

Staff

  • Photo
    Prof.
    K. KOMATANI
  • Photo
    Assoc. Prof.
    K. KOZAKI
  • Photo
    Assis. Prof.
    R. TAKEDA
  • Photo
    Assis. Prof.
    K. HAYASHI

Content of research

Intelligence of machines, e.g., to talk with humans, is still under development. To realize human-friendly and helpful robots, the spoken dialogue function is indispensable. We study basic technologies on spoken dialogue systems, such as acoustic signal processing, natural language processing, machine learning, knowledge representation and so on, across several layers from acoustic signal processing to social interaction. We have also been involved in the ontology engineering, which organizes human knowledge and describes it in a machine-readable format.

Current Research Topics

  1. 1. Active-listening dialogue systems adaptive to situations and users
  2. 2. Inferernce and understanding based on generalization of knowledge graphs
  3. 3. Domain knowledge acquisition through dialogues
  4. 4. Ontology engineering and Linked Open Data (LOD)
  5. 5. Sound source localization based on deep neural networks
  6. 6. Natural language processing using embeddings of large-scale knowledge graphs

Figure / Graph

  • Fig.1
    Spoken dialogue robots interacting with multiple participants

VIDEO INTRODUCTION