Rich Information Interaction in Digital Libraries

1
Rich Information Interaction in Digital Libraries
Ali Shiri Associate Professor School of Library
and Information Studies University of Alberta
2
Human-information interaction
Marchionini argues that the past 30 years
have seen a shift from distinct emphasis on
information, individual people, and specific
technologies to emphases on the interactions
among more diverse forms and amounts of
information, people and technologies.
Marchionini, G. (2008) Human-information
interaction research and development Library and
Information Science Research, 30(3), 165-174.
3
Rich interfaces

• Rich information, rich representation and rich
  interaction
• Many web-based applications and services
• are becoming informationally and visually rich
• support users exploration of the content
• Combine searching and browsing
• Terms used rich prospect interfaces, exploratory
  user interfaces, rich information interfaces,
  interactive user interfaces

4
Top 20 most hotly revised articles in Wikipedia
5
Rich Information and Rich Interactionin Search
Engines
6
Yahoo! Interface (December 15, 2008)
7
Visual search engines
8
(No Transcript)
9
Projects

• Thesaurus-enhanced visual interfaces
• Metadata-enhanced visual interfaces to digital
  libraries

10
Thesaurus-enhanced visual interfaces for
bilingual digital libraries
11
Why visual interfaces?

• Visual interfaces to digital libraries are
  becoming increasingly popular
• Information visualization techniques allow for
  rich representation of information within digital
  library interfaces
• They offer ways in which searching , browsing,
  exploration and navigation can be combined to
  provide a richer interactional experience

12
Why thesauri?

• A number of studies have found that
  thesaurus-enhanced search interfaces support
  users query formulation and expansion
• Searching, browsing and exploration of digital
  libraries can be enhanced using thesauri and
  taxonomies
• Commercial interfaces are increasingly using
  thesauri and other terminological tools as part
  of their search and browsing systems

13
(No Transcript)
14
A Study of Knowledge Organization Systems in
Canadian Digital Collections (2005)
15
SFU Library Editorial Cartoons Collection
16
Objectives

• To explore the ways in which subject access can
  be provided based on thesauri as an integral part
  of the search process.
• To explore the ways in which seamless access can
  be provided to the searching, browsing and
  navigation features within a thesaurus-enhanced
  interface

17
Theoretical framework

• The idea of rich prospect interfaces in which
  multiple representations can be used to allow
  access to a digital collection (Ruecker and
  Liepert, 2004).
• The interface should provide features and
  facilities to support thesaurus browsing, query
  formulation and results examination as seamlessly
  as possible (Shiri et al., 2002)

18
Choice of thesaurus

• The Government of Canada Core Subject Thesaurus
  was selected for this project because
• It is a bilingual thesaurus
• It is associated with a number of digital
  collections such as the Government of Canada
  Publications and the Government of Canada site
• It is a general purpose thesaurus
• It represents a standard thesaurus with term
  relationships
• It is a faceted thesaurus which makes it suitable
  for browsing purposes

19
Search box
Browse subject categories
Government of Canada Core Subject Thesaurus
interface
20
High level facets
Government of Canada Core Subject Thesaurus
browsing interface
21
Thesaurus terms
Government of Canada Core Subject Thesaurus Terms
22
The interface combines three spaces

• Query space for formulating search statements
• Thesaurus space for browsing and navigating the
  thesaurus
• Document space for viewing document
  representations

23
Document space
SearchLing user interface
24
Retrieved documents
25
User evaluation

• 15 participants
• Professors
• Graduate students
• Data gathering
• Morae
• Interviews
• Questionnaire (Usability)
• Tasks
• selection of single terms from thesaurus
• Combining terms from thesaurus
• Language sort feature
• Users own search topics

26
Findings

• Searching vs. browsing all but two of the users
  said that it was useful to have searching and
  browsing together
• Browsing was found more useful by users with no
  prior knowledge of the topics or for new research
  topics
• The bilingual users agreed that Searchlings
  ability to facilitate searches simultaneously in
  both languages is very useful

27
Findings (cont.)

• Users were not interested in using the interface
  as a translation tool, rather they would like it
  as a retrieval tool
• Query formulation using the thesaurus was found
  very useful
• Narrowing and broadening of the search and the
  number of documents
• The usefulness of the Search term pool

28
Metadata-enhanced Visual Interfacesto Digital
Libraries
29
Objectives

• 21 metadata-enhanced visual interfaces, developed
  in the 1990s and 2000s, were evaluated in terms
  of
• information access and retrieval features
• metadata elements
• visualization techniques and metaphors

30
(No Transcript)
31
Theoretical framework and methodology

• Information and computer sciences
• Metadata
• Digital libraries
• Human computer interaction
  Visual user
  interfaces
  Interface design
  
  Visualization techniques
  Visual
  metaphors
  
• Information retrieval
• Query formulation reformulation
• Relevance
• Interactive information retrieval

• Evaluation measures were developed based on
• Borner and Chens visual interfaces to digital
  libraries (2002) and
• Shneidermans Information visualization taxonomy
  

32
Table 1. Measures developed based on
Borner and Chens (2002) usage scenarios
Table2. Taxonomy of visualization techniques by
Shneiderman
33
Findings

• Increasing use of 3D and multi-dimensional
  visualization
• A noticeable shift towards the use of multiple
  visualization techniques and metaphors
• Increasing use of metadata in interfaces
  developed in the 2000s
• 15 interfaces support query construction or
  modification features.
• 19 interfaces offer search result visualization
  features.
• 14 interfaces provide visual collection
  representation facilities
• Only 8 interfaces have incorporated relevance
  visualization facilities

34
Design implications and further research

• Combining metadata and a subset of the collection
  on the interface as context.
• Visualizing item-level and collection-level
  metadata for multimedia digital collections
• User evaluation of metadata-based visual
  interfaces to support query formulation,
  reformulation, and result presentation.
• Design and evaluation of task-based and
  genre-based visual interfaces enhanced with
  metadata to support exploratory interaction