IRiX

Information Retrieval in Context

1. Introduction
2. Taxonomy of context features
3. Situation
4. Acquisition of context information
5. Aspects of information retrieval process to improve/support
6. Evaluation of context features
7. Types of IR/Search applications
8. Links
9. References

Introduction

This is a small attempt to understand and (more importantly) share the ideas of context that can be exploited in IR. The recent IRiX workshops suggest that there is ample room for investigating this emerging notion. Defining what are the context features is one thing, capturing the features accurately is another. Leveraging the features to improve parts of information retrieval processes and evaluating the effectivness appropriately are even more challenging issues. Providing an overview of context is the primary aim of this (wiki) page.

Another prospect of IRiX is that will facilitate closer communication between IR and related fields including Human Computer Interaction, Information Visualisation, and Information Seeking Research. The work such as IRiX/References#Kelly05 and IRiX/References#Larsen05 are a good example of successful application of concepts drawn from a cognitive model to improve the retrieval effectiveness. A number of models in IR and Information Seeking (or Behaviour) Research is discussed in IRiX/References#Wilson99. It would be fascinating if we could bring more of such studies into the field of IR.

This (wiki) page has been created as a part of the Adaptive Search Models for Information Retrieval Project. Feel free to contact me for your feedback. It's still in an early stage, but any suggestion for including related work is welcome.

Taxonomy of context features

(See IRiX/References#Ingwersen05b)

• Work or daily-life task or interest features

• Searcher features

• Interaction features

• System features

• Document features

• Environmental/physical features

• Temporal features

Situation

• Personal cognitive construct made in contexts (IRiX/References#Ingwersen05a)

  • Knowledge gap or ASK and relevance

  • Uncertainty and other emotional state

  • Potential sources for the solution of the work task or interest

  • Intentionality (i.e., goals, purposes, motivations, etc.)

  • Information preference, strategies, pressures, constraints (costs, time, etc.)

  • Self (i.e., own capabilities, health, experiences)

  • Systemic and interactive features and information object

• Best conceived as the totality of the values held by the facet variables such as (IRiX/References#Ingwersen05b):

  • users (/TaskFeatures and /SearcherFeatures);

  • systems (/SystemFeatures and /DocumentFeatures); and

  • environment (/EnvironmentalFeatures and /TemporalFeatures).

Acquisition of context information

• Explicit measures

• Implicit measures

• Integrated measures

Aspects of information retrieval process to improve/support

• Ranking

  • Re-ranking

  • Clustering

• Query re/formulation

• Interface/Visualisation

  • Query interface

  • Search results presentation

• Learning/Exploring

Evaluation of context features

• System-centred methodology

• User-centred methodology

• Simulated evaluation

Types of IR/Search applications

• Web search

• Personalised search

• Enterprise search

• Mobile search

• Geographic search

• Distributed search

• P2P Search

• Meta Search

• Email Search

• Collaborative Search

• Text retrieval

• Image retrieval

• Music retrieval

• Movie retrieval

• Spoken document retrieval

• XML retrieval

• Cross language retrieval

Links

• Conferences/Workshops

  • IRiX Workshop 2004

  • IRiX Workshop 2005

  • Symposium on Informaiton Interaction in Context (IIiX)

• Special topic issues

  • Special Issue on Journal of Information Retrieval (Contextual Information Retrieval)

  • Special Issue on Information Processing & Management (User-centred Evaluation of Information Retrieval Systems)

References

IRiX/References