LibStat a unified library usage statistics database

1
LibStat - a unified library usage statistics
database

Roxana Maria Popistasu Library systems
department Library of the University of
Amsterdam R.M.Popistasu_at_uva.nl
2
LibStat

• a new statistics tool for the electronic library
  systems used at the University of Amsterdam
• Goal gather statistical information that covers
  both quantitative and qualitative fields,
  providing reliable statistics on the total
  combined usage of databases and e-journals from
  all sources.

3
Implementation

• Dedicated database (MySQL)
• Sources of information
• log files (MetaLib, Apache etc.)
• databases of the library systems (MetaLib, SFX
  etc.)
• external sources (information from publishers and
  suppliers)
• Similar structure between systems gt make
  connections and extract simultaneously
  information from several systems
• No front end / user interface, but make the
  database accessible to third party reporting tools

4
Implementation

• Quantitative reports
• number of searches per periode / IP-range
• top 50 accessed journals
• etc.
• Qualitative reports
• navigation patterns (how far does a user get?)
• session properties
• Connections between systems
• for example, from MetaLib to SFX to Fulltext

5
Implementation
Metalib
MySQL Admin tool
Perl script
LibStat
SFX
MySQL export
MS Access (via ODBC)
ScholarlyStats
SUSHI
Navicat
Reports publishers
script / MySQL import
Other tools
Other systems
6
MetaLib sessions tables

• MetaLib reports
• not enough information for the needs of the
  University of Amsterdam
• users who have access to statistics shouldn't
  have administrator rights
• Sources for MetaLib statistical information
• Apache log files (the main source)
• web server log files (www_server)
• MetaLib Oracle database

7
MetaLib sessions tables

• Structure focused on the session entity

8
(No Transcript)
9
MetaLib sessions tables

• Special designed table state levels that can
  be reached during a session (patterns of
  navigation)

10
(No Transcript)
11
SFX sessions tables

• Very different situation from MetaLib
• the built-in SFX statistics are in a higher
  number and cover all aspects
• SFX MySql database contains thorough and clear
  information
• Why new statistics?
• overview information of all the systems
• easier connections between systems

12
SFX sessions database

• Source of information
• SFX MySql database mainly two tables
  STAT_OBJECT_OFFLINE and STAT_TARGET_SERVICE_OFFLIN
  E
• Structure
• similar to MetaLib, keeping the session as a
  central entity
• additional information for each step of the search

13
SFX sessions database

• Structure

14
(No Transcript)
15
Information from external sources

• Challenges
• different ways of presenting the information (not
  always complying with international standards)
• different formats for COUNTER reports
• different ways of delivering the information to
  the library (through a portal / by email etc.)
• reports for the entire University of Amsterdam,
  not for specific IPs (except of Elsevier)
• unclear information when it comes to sources
  (whether the search actions occurred only within
  the website / database of the publisher or they
  came from indirect link from other sources, like
  MetaLib or third party websites)

16
Information from external sources

• Structure (compliant with COUNTER reports)

17
(No Transcript)
18
Information from external sources

• Implementation
• Gather reports and statistical data from
  publishers and suppliers
• Put information into a standard format
• ScholarlyStats (www.scholarlystats.com) -
  collects and consolidates journal, database and
  e-book usage statistics from library vendors and
  delivers back COUNTER reports with information
  from all sources

19
Information from external sources

• ScholarlyStats strong points
• information from all vendors (regardless if it is
  COUNTER compliant or not)
• final COUNTER reports in both CSV files and Excel
  spreadsheets
• automatic data transfer (using the standard SUSHI
  protocol)
• easy and straightforward administration of the
  ScholarlyStats account
• possibility of archiving the usage statistics
  (both final reports and raw vendor source usage
  reports)

20
Information from external sources

• ScholarlyStats weak(er) points
• reports are only run once a month
• user information is not available (using COUNTER
  reports)
• ScholarlyStats just harvests the information it
  does not modify it in any way
• data only as good as the provider presents it
• searches within the website / database of the
  publisher or from other sources?

21
Using the database
MySQL Administrator
22
(No Transcript)
23
Using the database
MySQL Query Browser
24
(No Transcript)
25
Using the database
MS Access (via ODBC)
26
(No Transcript)
27
(No Transcript)
28
(No Transcript)
29
Using the database
Navicat MySQL Admin
30
(No Transcript)
31
(No Transcript)
32
Using the database
Navicat Reports
33
(No Transcript)
34
(No Transcript)
35
(No Transcript)
36
(No Transcript)
37
Further steps

• investigate options for adding information from
  other library systems (ILS etc.)
• add information from other external sources
• create a tool for archiving and reloading
  statistical data (like in SFX)
• create procedures for automatically updating the
  databases from all the sources
• investigate options for cooperation with other
  libraries

38
Questions?
Thank you for your attention!

• Further information
• Roxana Maria Popistasu R.M.Popistasu_at_uva.nl
• Lukas Koster L.Koster_at_uva.nl