Pesch, NISO Update: IOTA
0 likes•84 views
The document outlines recent and current activities related to the NISO initiative focused on enhancing OpenURL quality through analytics and addressing challenges in SUSHI reporting protocols. It discusses the development of a mathematical formula for scoring OpenURLs to predict link resolution success, and details the formation of a committee tasked with improving standards in usage statistics harvesting. Additionally, it highlights efforts to simplify SUSHI client testing and enhance automation within usage reporting while maintaining transparency and industry-wide metrics.
Pesch, NISO Update: IOTA
- 1. NISO Update January 22, 2012 Oliver Pesch Chief Strategist, E-Resources EBSCO Information Services
- 2. • Quick review • Recent activities • Current activities
- 3. • IOTA stands for: Improving OpenURL Through Analytics • Offshoot of work done by Adam Chandler trying to answer the question: Why do some OpenURLs not resolve?
- 4. The theory • If you know what data elements are required to create target links • You should be able to predict the probability of success for any given OpenURL based on the elements included in the link
- 5. The Completeness Score • A mathematical formula that quantifies the probability of success by scoring the OpenURL based on the data elements included, where the value of each element is weighted based on how often it is needed in a target link • Group all the OpenURLs for a given Source and look at their scores in aggregate – you can rate the provider of the OpenURL
- 6. • Committee Members o Adam Chandler, Cornell University (Chair) o Aron Wolf, Serials Solutions o Clara Ruttenberg, University of Maryland o Elizabeth Winter, Georgia Tech o James Wismer, Thomson Reuters o Oliver Pesch, EBSCO Information Service o Rafal Kasprowski, Rice University o Susan Marcin, Columbia University • Charge o Investigate the feasibility of creating industry-wide, transparent and scalable metrics for evaluating and comparing the quality of OpenURLs
- 7. • OpenURL Quality Web Site o Processed over 21.5 million OpenURLs o Calculated Completeness Scores o Provides Completeness Scores by SOURCE (OpenURL Providers) o Provides metrics on elements frequencies http://openurlquality.niso.org
- 8. • Conducting validation experiments o Confirm a correlation between the Completeness Score and the OpenURLs ability to create an actual full text or ILL link. o EBSCO and Serials Solutions participating o Using a sample size of 10s of thousands of OpenURLs
- 9. • Complete validation experiments • Adjust element weights, if necessary • Complete a final report for the committees work with recommendations for further work • Continue to maintain the OpenURL Quality web site as a tool for librarians and content providers
- 10. NISO Update January 22, 2012 Oliver Pesch Chief Strategist, E-Resources EBSCO Information Services
- 11. • Quick review • Current activities o Support for Release 4 of the COUNTER Code of Practice o SUSHI Server Test Mode Recommended Practice o SUSHI Server Status Report o COUNTER SUSHI Implementation Profile
- 12. • Standardized Usage Statistics Harvesting Initiative • Z39.93 (2007) • Addresses the labor intensive and repetitive task of retrieving COUNTER reports
- 13. • Committee Members o Bob McQuillan, Innovative Interfaces, Co Chair o Chan Li, California Digital Library o John Milligan, ScholarlyIQ o Paul Needham, Cranfield University o Oliver Pesch, EBSCO Information Services, Co-chair • Charge o Standing committee responsible for the SUSHI standard o Maintenance Agency for the COUNTER XML Schema
- 14. Addressing… • New formats • New devices used for access • New access technologies • Need for automation • Need for machine processing
- 15. Changes from prior release… • New reports • New metric types • New data elements What this means for the SUSHI committee • Update the COUNTER XML Schema • Update the COUNTER Enumeration Schema • Update the reference documents on the web site
- 16. The problem • Developers find it difficult to test their SUSHI client against production versions of SUSHI Servers o Complicated requirements for authentication o Limits to the number of requests per day o Test data may not represent reality The solution • A recommended practice o Lowers the barrier for the developer o Simplified registration o Removed authentication hurdles o Allows multiple re-tries in a day o Data representative of live data
- 17. The goal • A simple report that can be requested via SUSHI • Provides status information o Is the server able to deliver usage o What dates of usage are available • Provides configuration information o Lists COUNTER reports that are supported o Instructions for configuring a client o Authentication information o Day of the month when usage available
- 18. Why • SUSHI and COUNTER tend to be abstract • Abstraction leaves the opportunity for concepts to be interpreted differently • Different interpretations leads to inconsistencies • Examples: o Different authentication methods that require custom development o Wrong capitalization on controlled data values o Including “Totals” in the XML o Including invalid ISSNs (e.g. “N/A”)
- 19. Authentication for SUSHI Client • IP and use of requester and customer IDs allowed • Use of SOAP extensions not allowed
- 20. Report Names • Must exactly match the value in the “Name” column on the SUSHI Reports Registry Date Ranges for SUSHI Request • Format must by yyyy-mm-dd • Begin date must be first of the month • End date must be last day of the month • End date must be later than begin date
- 21. SUSHI Exceptions • Standard error numbers must be used for standard error conditions (see Table 17 in the SUSHI standard) SUSHI Server Registry • SUSHI Server must be registered on SUSHI Server Registry
- 22. Data to Return in the COUNTER Report • Only the months of usage asked for • Include exception if some or all usage not available • Only include items appropriate to the report (e.g. don’t return book usage in a Journal Report) • Do not include totals for the report or an item
- 23. Data Elements Values • Use registered values with proper casing for enumerated values (Item Data Type, Category, Metric Type, etc.) • Use values appropriate for the report
- 24. Designed for use by implementers • Prescriptive in nature • Lots of tables and charts
- 25. Designed for use by implementers • Prescriptive in nature • Lots of tables and charts
- 26. • Finalize and publish… o COUNTER SUSHI Implementation Profile o SUSHI Server Status Report o SUSHI Server Test Mode Recommended Practice • Continue our work in supporting developers and librarians • Continued promotion and education