Context aware interactive information retrieval
Download
1 / 23

Context Aware Interactive Information Retrieval - PowerPoint PPT Presentation


  • 120 Views
  • Uploaded on

Context Aware Interactive Information Retrieval. Claus-Peter Klas Paul Landwich Matthias Hemmje Dagstuhl , Seminar on Interactive Information Retrieval,, March 2009 . Outline. User / Use Cases / User Scenarios Context Information Dialog Daffodil / Framework / Logging

loader
I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.
capcha
Download Presentation

PowerPoint Slideshow about ' Context Aware Interactive Information Retrieval' - damia


An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.


- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript
Context aware interactive information retrieval

Context Aware Interactive Information Retrieval

Claus-Peter Klas

Paul Landwich

Matthias Hemmje

Dagstuhl, Seminar on Interactive Information Retrieval,, March 2009


Outline
Outline

  • User / Use Cases / User Scenarios

  • Context

  • Information Dialog

  • Daffodil / Framework / Logging

  • Relevance Feedback

  • Summary & Outlook


User scenarios
User Scenarios

  • “Easy” task: Known-Item-Search

    • Supported by Google and libraries

  • Complex task: Thesis, Related work chapter, etc.

    • User has incomplete or no knowledge about domain

    • User has partial knowledge about domain


Context what is it and how to gain it
ContextWhat is it and how to gain it?

  • Consider a computer scientist (e.g. thesis)

    • General

      • Affiliation, research domain, position, list ofpublications, geospatialdata

    • Task

      • Articles

      • Conferences

      • Journals

      • Authors

    • Historic interactions with system(s)


Approach to iir information dialog model
Approach to IIR:Information Dialog Model

  • What are possible activities?

  • How do these activities affect our information dialogue context?

Information Dialogue Context



Activities
Activities

visualised result set

  • Exploration

  • Navigation

  • Focus

  • Inspection

  • Evaluation

  • Store

I: Content set

J: Interest set

R: Relevance set

r: Result set

k: Recall set



D affodil framework interactive ir
DAFFODIL FrameworkInteractive IR


Scientific work or information journey

Create knowledge

re-present

re-present

retrieve

Search knowledge

interprete

interprete

Interpret

results

collate

collate

Structured storage

Scientific Work orInformation Journey

Choose information sources

discover


Rich interaction levels of logging
Rich interaction: levels of logging

  • User behavior

  • User-System interaction

    • Conceptional events

    • Service events

    • Interface action events

    • UI events

  • System parameters


User behavior
User behavior

  • User behavior can never be captured completely, because it is inside the users head.

  • What we can do is:

    • Captured behaviour by supervision

    • Video

    • Questionare


User system interaction
User-System Interaction

  • Here we have plenty to capture, although, only interpretation of the data without user behavior is tricky.

  • Events

    • Conceptual events: High level abstract events for comparison

    • Service events: Events special for a IR service

    • Interface action events: Events from entering data, click buttons, select menue

    • UI events: Events from keystrokes, mouse movement


System parameters
System parameters

  • On this level all information about the soft and hardware is captured.

  • Parameters

    • Software: Efficiency of algorithms

    • Hardware: Usage of computer resources

    • Network: Load of computer network


Conceptual events
Conceptual Events

  • Search

  • Inspect

  • Browse

  • Annotate

  • Help

  • Navigate

  • Display

  • Store

  • Author

  • Communicate



Relevance feedback
Relevance Feedback

  • Investigate two RF scenarios for query term suggestion and re-ranking the result list with different approaches.

  • Use result list (current situation) and task path events (> 1 session)



Evaluation
Evaluation

  • Investigate on methods and models to interpret log data

  • Evaluate the relevance feedback and show the user the difference and may ask for explicit feedback


Summary outlook
Summary & Outlook

  • Model for capturing the information dialog

  • Daffodil Framework

    • Log data over session bounderies (xx GB)

    • Running „real live“ testbed in computer science (other)

    • Manysophisticatedfunctionalities

  • Examplefor IIR: Relevance Feedback

  • General: Daffodilframeworkcanbe a evaluationframeworkfor IIR

  • Try It Out: http://www.daffodil.de


Goal cognitive enhanced model of information retrieval
Goal: Cognitive enhanced model of information retrieval

Problem

Information

deficit

Information

need

Query

Adjustment

Discovery

Presented

knowledge

Cognition

1. State

(concrete)

2. State

(uncertain)

3. State

(fuzzy)

Knowledge

Stored

knowledge

Represented

knowledge

Core IR-engine

Cognitive enhanced IR-User interface

Human

[Lan07]


Statistics history of d affodil
Statistics/History of DAFFODIL

  • DAFFODIL started in 2000 as national funded project @ University of Dortmund in the IR group of Norbert Fuhr

  • 2 PhDs, more to come, > 14 Master/Bachelor thesis,

  • > 14 Publications in JCDL, ECDL, etc.

  • Lives unfunded in teaching, projects and as evaluation framework now at Duisburg-Essen and Distance University of Hagen

  • Projects: Daffodil, DELOS NoE on DL, SHAMAN (EU), EMTO (Philosophers), LACOSTIR, INEX


Adaptive framework concepts
Adaptive Framework & Concepts

User model

Personalisation

Recommendation

Adaptivity

Awareness

Kollaboration


ad