Faceted Maps of Knowledge and Domains: Peeling the Onion of an Idea1, An Examination of the use of F

1
(No Transcript)
2
Faceted Maps of Knowledge and Domains
Peeling the Onion of an Idea1,
An Examination of the use of
Facet Analysis in Website Design(1) In
Calvin Mooers words, Facet provides an
analytical tool that is, the idea of facet
allows you to peel the onion of an idea. Brian
Vickery. Faceted Classification Schemes. In Susan
Artandi (Ed.) Rutgers Series on Systems for the
Intellectual Organization of Information. V. 5.
New Brunswick, NJ Graduate School of Library
Science at Rutgers University
3
Research Overview

• Exploratory
• Current use of facet analysis (FA) and faceted
  classification (FC) for display, and searching on
  the web.
• Applications of FA in website design

4
Research questions

• How has Facet Analysis been used in the
  architecture of websites?
• Does the use of Facet Analysis mark a shift in
  the base structure of websites?
• Might the current use of Facet Analysis on the
  Web demonstrate its potential for Information
  Retrieval and other areas?

5
Outline

• Situating the research
• Knowledge Organization and Information Retrieval
• Facet Analysis and Faceted Classification
• Research overview
• Research questions
• Methodology
• Trajectory

6
Domain Analysis
(FA) Facet Analysis
WEB
Knowledge Organization
Information Retrieval
FA
Information Architecture Knowledge Management
INFORMATION SCIENCE
7
FC FA
Knowledge Organization
Information Retrieval
(FA) - Facet Analysis (FC) Faceted
Classification
INFORMATION SCIENCE
8
Facet Analysis (FA)Faceted classification (FC)

• FA - (analytical technique)
• Listing of characteristics of the entities in a
  universe (exhaustive, mutually exclusive)
• FC - (synthetic structure)
• Division of entities in a universe (by one
  characteristic at a time) FC (structure of
  synthesis)
• Synthesis combine relevant facets
• Schedule of terms for description
• Assignment of notation

9
Background FA/FC

• Universal Decimal Classification
• Otlet, La Fontaine -Documentalists
• 1904-1907 scheme published
• Bliss Bibiliographic Classification
• Henry Evelyn Bliss
• 1908 (practice) 1923-1933 (theory)
• Colon Classification
• S. R. Ranganthan,
• 1933 (practice) 1937-1967 (theory)
• (La Barre, 2000, 2003)

10
Early workpractical/ theoretical

• Library Circle
• 1951 - Bangalore/New Delhi
• Classification Research Group
• 1952 - Britain
• Classification Research Study Group
• 1957 - North America
• (La Barre, 2004a, 2004b, 2004c)

11
Application/Library Documentation

12
Early work FA/FC

• 1957 International Conference on Classification
  for Information Retrieval Dorking,
  England.(CRG/ others) 
• 1958 English Electric Scheme -- Bins
• (first of a series of faceted schemes by CRG
  members)
• 1962 ASLIB-Cranfield test Faceted scheme for
  aeronautics.
• -- Cleverdon
• 1963- American Institute of Physics AUDACIOUS
  (UDC for IR). -- Atherton Cochrane/
  Freeman
• 1964 Second International Study Conference for
  Classification Research, Elsinore, Copenhagen-
  Atherton Cochrane (Ed.)
• 1965 Western Reserve University (CDCR)
• Aslib-Cranfield test Metallurgical documents
  (using English Electric Scheme) -- Rees/
  Aitchison/ Cleverdon
• Semantic code factoring procedure for IR
  (influenced by CC)
  -- Melton/ Kent/ Perry
• 1965 - American Meteorological Society.
  Mechanization of UDC for retrieval -- Freeman/
  Rigby

13
Revision CC/BC2, Criticism, IR applications,
Automation, Software reuse repositories
Thesauri (Thesaurofacet/NASA/UNESCO) Extension
BC2/ CC, BSO, PRECIS, POPSI
1972
14
(FA) Facet Analysis
WEB
Knowledge Organization
Information Retrieval
FA
Information Architecture Knowledge Management
INFORMATION SCIENCE
15
Metadata, Database theory, Image archives,
AAT(1990), Organizing the Web.
IR applications Faceted search, browse and
retrieval, Web design IA and KM
16
Commentary

• Faceted classification is one of the most
  powerful, yet least understood, methods of
  organizing information.
• Peter Merholz Innovations in Classification.
  (September 2001) http//www.peterme.com/archives/0
  0000063.html

I personally find the term facet to be
confusing. I prefer the terms attributes and
attribute values. These terms are used in both
the database world and the artificial
intelligence world, to describe a very similar
functionality, sometimes the exact same
functionality. ReSIGIA-l Faceted approach
applied to content From Donna M.
FritzscheDate Fri Nov 14 2003 - 135423 EST
http//www.info-arch.org/lists/si
gia-l/0311/0161.html
My complaint is that there is a lot of talk about
facets, but little of any substance. Most of it
won't help you build your own faceted
classification scheme. It amounts to saying the
grass is greener on the other (faceted) side, but
fails to give you a map explaining how to get
there and what obstacles you'll face along the
way. And the academic literature doesn't help
much either. It's too dense and I can't recommend
it to the practitioner (not the stuff I've seen).
June 22, 2002 Christina Wodtke
http//www.eleganthack.com/MT/mt-tb.cgi/2
Faceted classification serves up multiplepure
classification schemes rather than a single
motley Taxonomy. Rosenfeld, L Morville, P.
(2002). Information Architecture for the World
Wide Web. 2nd Ed. Cambridge, MA OReilly.
17
Timeline

• September 2001 - Peter Merholz
• Innovation in classification
• January 2002 - Travis Wilson FacetMap
• March 2002 Phil Murray KMConnection
• April 2002 - Discussion on SIG/IA list
• May 2002 Peter Van Dijck XFML
• July 2002 Rosenfeld and Morville Information
  Architecture for the World Wide Web. 2nd Ed.
• December 2002 Van Dijck/ Murray establish
• Faceted Classification discussion list

18
Facet Analysis (FA)Faceted classification (FC)

• FA - (analytical technique)
• Listing of characteristics of the entities in a
  universe (exhaustive, mutually exclusive)
• FC - (synthetic structure)
• Division of entities in a universe (by one
  characteristic at a time) FC (structure of
  synthesis)
• Synthesis combine relevant facets
• Schedule of terms for description
• Assignment of notation

19
More FA / steps

• Identify domain / entities
• Mapping the scope
• (Context) Examine the domain
• (Content)Survey the literature
• (Users) Who? Information needs?
• Label/ sort
• Begin analysis with a list of standard
  categories (provisional guide) PMEST/ Who/
  Where/ How/ What/ When
• Result set of homogeneous mutually exclusive
  groups (facets)
• Formulate every distinctive logical category and
  possible relation
• Cluster /order
• In-depth analysis of categories
• Cluster terms/ objects into arrays or groups
  which share a common characteristic

20
Buildings
21
Facet analysis

• Fundamental concepts are analyzed and
  grouped together as facets
• Hunter, E. (2002) Classification made simple.
  Ashgate
• Building Facets
• Location
• Composition
• Purpose
• Date/Period constructed
• Performance
• Style
• Associated persons
• ETC. . .

22
Faceted search
421,291 images
23
Faceted search
421,291 images
24
(No Transcript)
25
1,050 images
26
163 images
27
15 images
28
421,291
29
Faceted navigation
Sent vrijdag 9 januari 2004 013 To
facetedclassification_at_yahoogroups.com I'm
Pete Bell from Endeca, a software company based
in Cambridge, MA Since 1999, we've been
developing Search and Guided Navigation software
that makes it easier for people to find
information What is Guided Navigation?
Endeca's brand name for our flavor of faceted
search and navigation.
30
(No Transcript)
31
CompUSA
32
Outline

• Situating the research
• Knowledge Organization and Information Retrieval
• Facet Analysis and Faceted Classification
• Research overview
• Research questions
• Methodology
• Trajectory

33
Research Overview

• Exploratory
• Current use of facet analysis (FA) for display
  and organization of entities on the web.
• Applications of FA in website design

34
Research questions

• How has FA been used in the architecture of
  websites?
• Does the use of FA mark a shift in the base
  structure of websites?
• Might the current use of FA on the web
  demonstrate its potential for IR and other areas?

35
Conceptual FrameworkDomain Analysis - HjØrland

• Socio-cognitive perspective the world is
  physically, socially and subjectively
  constructed.
• Knowledge is explicit (communicable through
  language) and implicit (embedded within
  activity).
• Identify, describe, organize and communicate
  information resources in order to serve specific
  goals.
• Acknowledge the importance of historical, social,
  cultural contexts and their influence on
  cognitive processing.
• Role of discourse communities (common goals,
  interests, purposes).

36
Research trajectory

• Website survey
• Practitioner survey and interviews
• Primer construction
• Guidelines for assessment
• Use, impact, future of facet analytical theory
• for applications
• de novo theoretical

37
Methods

• The chosen methods should flow from the research
  questions, not vice versa. Alford (1998, p.1).
• Caveat
• Potentially useful methods
• Comparative (general) discussion

38
Potential Methods

• Content analysis
• Survey
• Interviews
• Usability testing
• Transaction log analysis

39
General issues

• Triangulation
• Multiple sources/ methods/ investigators/
  theories
• Web concerns
• Sampling (representative?)
• Persistence (of sites)
• Coding frame (timing/ archiving)
• Reliability
• Results statistically significant
• Repeatability
• Generalizability
• Can findings be applied appropriately?
• Increased by attention to reliability and
  validity
• Validity
• Does study measure what it intended?
• Is design and measurement rigorous?

40
Content Analysis - Issues

• Sampling
• (units)
• Reliability
• (of coding categories)
• Generalizability
• (to theory)
• Validity
• (of inferences)

41
Survey - Issues

• Sampling
• (unit of analysis/ frame)
• Generalizability
• (representativeness)
• Reliability
• (generally high)
• Validity
• (problematic)

42
Interviews - Issues

• Generalizability
• (limited)
• Reliability
• (general interview plan/ probes)
• Facts vs Memory
• Validity
• (compare with written)

43
Usability Testing - Issues

• Sampling
• (users/ websites)
• Generalizability
• (representativeness of task/ users)
• Reliability
• (proper sample size/ user differences)
• Validity
• (confounds/controls/ representative users and
  tasks)

44
Transaction log analysis - Issues

• Limited to search logs
• Difficult to separate human/machine searches
• Effect of caching incomplete logs
• Reliability and validity affected

45
Conclusion

• Potential
• Capture current practices on the web
• Implications for web designers
• Practice to inform theory
• Theory to inform practice

46
References

• Hunter, Eric. (2002). Classification made simple.
  Gower/Ashgate.
• Mills, J. (2004). Faceted classification and
  logical division in information retrieval.
  Library Trends 52(3), 541-570.
• Ranganathan, S. R. (1959). Elements of library
  classification. London Association of Assistant
  Librarians.
• Ranganathan, S. R. (1937/1957/ 1967). Prolegomena
  to library science. New York Asia Publishing.
• Vickery, B. (1960). Faceted classification A
  guide to construction and use of special schemes.
  London Aslib.
• Sites
• CompUSA http//www.compusa.com
• Endeca http//endeca.com
• Images of England http//www.imagesofengland.org.
  uk

47
Questions?
48
Commentary
49
Peter Van Dijcks Guide to Ease (blog) Comments
on the Bliss Classification Association
Bibliographic Classification Guide. 10
October, 2002. http//www.poorbuthappy.com/ease/ar
chives/001603.html

• The standard categories recognized in
  classical facet analysis are Thing kind
  part property material process operation
  patient product by product agent space
  time. ltBCA websitegt
• Why do these people try to specify what facets
  are useful? Surely if I want to classify my
  content using a facet called How impressed I was
  with this content when I first saw it then that
  is a perfectly valid facet? This underlying
  assumption that there are generic or correct
  ways of categorizing the world is just, well,
  wrong. The world doesnt have an inherent
  classification that we just have to discover. All
  classification gets its meaning from the people
  using it, not from the objects being classified.

50
Travis Wilson (creator of FacetMap) Thread
Special vs. General Schemes Date Wed, 12 Mar
2003 091411 0000 http//article.gmane.org/gmane
.comp.infodesign.facetedclassification/170/matchf
acet

• At the other extreme, the universal qualities
  espoused by Ranganathan and others just don't
  apply to many resources that deserve
  classification. The "space" facet only works in
  the context of resources that have physical
  incarnations in a spatial location perhaps this
  is obvious, but it confirms that there is indeed
  a context -- one that many data-based resources
  don't share.
• In the end, the whole idea of a set of general
  facets as a "framework for facets typical of each
  discipline" simply limits the number of
  disciplines that can use FC (to those that fit
  within the framework). Ranganathan was really not
  concerned with disciplines outside of biology. I
  don't think that kind of limitation is
  appropriate for faceted classification in
  general, and I hope we don't impose it.

51
Example Literature Hierarchical/ Faceted

• English
• Prose
• Poetry
• sonnet
• ballad
• French
• Prose
• Poetry
• sonnet
• ballad
• German
• Prose
• Poetry
• sonnet
• ballad

• Language
• EN
• FR
• GR
• Form
• Prose
• Poetry
• sonnet
• ballad

52
Illustration of faceted and hierarchical website
structures

• Rosenfeld, L. Morville, P. (2002). Information
  architecture for the World Wide Web. Cambridge,
  MA OReill. P. 205.

53
Definition

54
Facet

• Facet
• A generic term used to denote any component of a
  compound subject, ltincludinggt ranked forms, terms
  and numbers (Ranganathan, 1967, p. 88).
• A general manifestation of a subject
  (Ranganathan, 1962, p. 82).
• Groups of terms derived by taking each term and
  defining it, per genus et differentiam, with
  respect for its parent class (Vickery, 1960, p.
  12).
• Facet analytical approach Proper and rigorous
  practice of facet analysis by observing the rules
  of logical division. (Broughton, 2001, p. 67
  Mills, 2004, p. 268).
• one characteristic of division is applied at a
  time conceptual analysis
• division steps should be logical and proximate
• division should be exhaustive (Mills, 2004, pp.
  551).

55
Planes of work

• Idea The work of FA takes place in the Idea
  plane, where an entity is analyzed into component
  parts
• Verbal FA continues here as further sorting and
  transformation of the selected categories/facets
  or terms occur.
• Notational work of FC -- translating selected
  terms into notation.

56
Facets continued

• Supports the notion that information can be
  assigned to multiple dimensions and may have
  individual attributes not intrinsic to the
  information.
• Small components of larger entities/units.
• Properties
• Attributes
• Characteristics
• Slots
• Relations
• Functions
• Concepts

57
High level categories
58
Classification

• The placing of subjects into categories
• The process of determining where an information
  package fits into a given hierarchy and then
  assigning the notation with the appropriate level
  of hierarchy to the information and to its
  surrogate.
• Taylor, A. (1999). The organization of
  information. Englewood, CO Libraries Unlimited.

59
Classification- assigning objects to classes

• Objects to be classified into a system are called
  elements, cases, units, exemplars, specimens or
  items. They are the sources or carriers' of
  properties, characteristics or variables.
• A property can only be useful in a
  classification, if it varies within the set of
  objects, that is, if at least two different
  values (categories, states, labels) on the
  respective property occur in the sample.
• When more than one property is used to
  characterize an object, the object can be
  described as a vector of values, a profile, a set
  of symptoms, or a pattern of features. The
  crucial assumption underlying classification is
  that objects are elements of a class, of a set,
  of a partition or-in biology-of a taxon. In other
  terminologies, the terms category' or cluster'
  are also used.
• Classification is the process of finding classes
  and of assigning entities to these classes. The
  end product of this order-creating process,
  however, is often also referred to as
  classification.' To stress this distinction, the
  term classification system' can be used for the
  end-product, although in clinical psychology and
  biology the word taxonomy' is more common.
  Identification is the assignment of a specific
  case or object to (usually only) one of the
  classes.
• H. Feger (2001). Classification Conceptions in
  the Social Sciences. International Encyclopedia
  of the Social and Behavioral Sciences, pp.
  1966-1973.