Design of Intelligent Human Interface Brief Summary of Research Activities - PowerPoint PPT Presentation

1 / 7
About This Presentation
Title:

Design of Intelligent Human Interface Brief Summary of Research Activities

Description:

(Currently: Tokyo Metropolitan Institute of Technology) PREST, Japan Science and Technology Corporation. 2 /16. Characteristics of VSA ... – PowerPoint PPT presentation

Number of Views:55
Avg rating:3.0/5.0
Slides: 8
Provided by: hrtDisT
Category:

less

Transcript and Presenter's Notes

Title: Design of Intelligent Human Interface Brief Summary of Research Activities


1
Design of Intelligent Human Interface- Brief
Summary of Research Activities -
  • Yasufumi Takama
  • Tokyo Institute of Technology
  • (Currently Tokyo Metropolitan Institute of
    Technology)
  • PREST, Japan Science and Technology Corporation

2
Characteristics of VSA
VSA 1
  • Multi-modal Interface
  • Face-to-face interaction speech dialog
  • Connected with WWW browser
  • Suitable for users with little knowledge of
    computers
  • Applicable to facility guidance systems
  • Dialog management system (DMS) with learning
    module
  • Acquisition of cooperative answering strategy

3
DMS with Learning Module
VSA 2
  • Cooperative answer depends on both users
    knowledge systems location
  • E.g. Where is the library?
  • Tell the way from systems location to the
    library
  • Use the landmark that user seems to know
  • 2 types of knowledge acquisition
  • Overlay-type user-model to estimate the knowledge
    state of the present user
  • Reinforcement learning to acquire
    location-specific appropriate rules

4
Document Ordering while Reading is an Effective
Way of Dealing with Vast Collection of Documents
FISH VIEW 1
Viewpoint-based support by Fisheye Matching
Making diagrams
Viewpoint extraction
Users Activity
Viewpoint information
feature generation
Reading documents
Retrieving new documents
5
Extracted Concepts as Viewpoints
FISH VIEW 2
  • From medical news
  • From cinema reviews
  • Categories of action, SF, monster, etc.

6
Query NetworkPlastic Clustering for
Visualization
Query Network 1
  • Clustering-based visualization applied to WWW-IR
  • Categorization (static, i.e. user independent)
  • Naive clustering (one-shot, i.e. no inheritance)
  • K-means, STC, etc.

Plastic clustering (gradual formation through a
series of IR)
7
Analogy to Immune Network
Query Network 2
  • Clusters generated for retrieved documents
  • Effective cluster can be reused for subsequent
    retrieval
  • Various clusters in plastic structure
  • Antibodies generated for antigens
  • Activated antibody can be sensitive to next
    invasion
  • Various antibodies in plastic structure

Activation value calculation based on
differential equation of mathematical biology
Write a Comment
User Comments (0)
About PowerShow.com