Cybermetrics Definitions and methods for an emerging discipline

1
CybermetricsDefinitions and methods for an
emerging discipline
Association pour la mesure des sciences et des
techniques

• Isidro F. Aguillo
• CINDOC-CSIC (Spain)
• isidro_at_cindoc.csic.es

2
Definition
3
Definition (II)
4
Definition (III)
5
Academic websize
6
New units
7
Rich files
8
Applying old methods

• Informetrics
• Data behaviour in the RD cyberspace
• Informetric distributions
• Dynamics and evolution in the Web and email fora
• Scientometrics
• Citation analysis
• Impact factor in the eWorld Electronic Journals
• RD evaluation Quality assessment
• Formal and informal communication in the
  cyberspace
• STM indicators
• Case studies
• International Cooperation
• Development countries, Gender, etc
• Bibliometrics
• Deep Web/ Invisible Internet

9
A new discipline (I)

• Traditional approach shortcomings Theory
• Websites are not papers
• But they can cover extra RD activities
• Web communication is informal
• But it is sometimes peer-reviewed or quality
  controlled
• Hypertext links are not Bibliographic citations
• But analysis methods could be the same
• Structure of the Web RD is chaotic
• But due to different information architectures
• Traditional approach shortcomings Methods
• Available databases (search engines) are not
  good
• No exhaustive coverage, but you can use several
  of them simultaneously
• Irregular behaviour, but you can use your own
  robots/agents

10
A new discipline (II)

• New research agenda Cyberinformetrics
• Web data architecture (KO, data mining)
• Hyperlink topology
• Comparative search engines analysis
• New research agenda Cyberscientometrics
• Non-publication oriented activity of scientists
• RD cybergeography and cyberdemography
• New units Institutional websites
• Real world and virtual ones
• New indicators
• Visibility (WebIF)
• Popularity Users behaviour
• New research agenda SciTechnoEconometrics
• Sitations analysis is better suited for uncover
  technological and economic impact of the research
  activity

11
What is needed?

• Theoretical developments
• A new sitation theory
• Explanations about and behind linking behaviour
• A classification of sitations
• Powerful tools
• Automatic data extraction software
• Reliable, flexible and comprehensive
• Automatic data classification software
• Data mining
• Improved visualisation techniques
• Empirical data
• Define and evaluate new statistics
• Build a full set of new sitation databases
• Including impact, visibility and popularity
  indicators

12
Broadening metrics research

• Cybermetrics/Webometrics is not only devoted to
  apply traditional scientometrics research agenda
  to the information in the Internet
• Although descriptive analysis of the RD output
  and scholarly communication in the Net are badly
  needed
• Although applying informetric analysis to the
  Web is theoretically very interesting
• Although eJournals and informal publication
  requires further analysis
• But Cybermetrics/Webometrics is also a complete
  new discipline
• With its own different theories to be built,
  tasks to be done, units to be defined, methods to
  be developed and problems to be solved

13
Description by agents and bots

• Automatic mapping
• Quantitative description of websites
• Compilation of the external links
• Visibility
• Web Impact Factor (WebIF)
• Search engines
• Altavista
• FAST (alltheweb)
• Hotbot
• (Google)
• Popularity
• Relative popularity (Alexa)

14
Mapping Agent
15
Alexa
16
Building descriptive indicators
17
Webindicators
18
Sampling EU RD in the Web

• Identification of the University websites
• Paper directories (World of Learning)
• Web directories and search engines
• Manual indexing
• Geographic codes (NUTS)
• Subject allocation
• UNESCO codes
• To be superseded by ISI classification
• Institutional level
• University, Faculty
• Department, research team or group

19
EU University sites (Jan. 2002)
20
Impact of EU webspace

• Web Sample
• Altavista (www.altavista.com)
• EU, G7, NAFTA, OECD, INT

21
EU and the OECD in the web

• Web Sample
• FAST (www.alltheweb.com)
• OECD 100 INT
• G7 (6030) INT
• NAFTA 60 INT
• EU 30 INT
• (incl. 100 eu.int)

22
Future developments

• Graph theory and the web
• How big is the graph? How many links on a page
  (outdegree)? How many links to a page (indegree)?
• Can one browse from any web page to any other?
  How many clicks?
• How different is browsing from a random walk?
• Can we exploit the structure of the web graph for
  searching and mining?
• What does the web graph reveal about social
  processes which result in its creation and
  dynamics?
• Is it possible to apply small world theory to
  the Web?

23
Time for demonstrations

• Recovering Infranet resources
• Clients Z39.50
• Size, density and link quality
• Link-checkers
• Quantitative description of the website
• Agents for downloading and mapping websites
• Search engines information recovery
• Agents for automatic recovery of data from search
  engines

24
Questions?
Thank you