IOTA: Improving OpenURLs Through Analy9cs. Where we re headed. Background 10/21/10

Transcription

1 IOTA: Improving OpenURLs Through Analy9cs Elizabeth Winter Adam Chandler October 13, 2010 Where we re headed Elizabeth Winter Adam Chandler NISO IOTA project Background Problem / Goals Benefits Current work Issues Upcoming events / Ways to follow IOTA Background OpenURL standard IOTA = Improving OpenURLs Through Analy9cs Working Group History: Adam Cornell L Année Philologique Mellon Grant NISO proposal 1

2 Poll #1 The Problem 72% of respondents to an online survey agreed/strongly agreed that a significant problem for link resolvers is the genera[on of incomplete or inaccurate OpenURLs by databases. Poll #2 Why don t OpenURLs always work? Poll #3 2

3 Goals Two- year research project Produce a technical report Develop a qualita[ve report to help OpenURL provider compare their OpenURL quality to that of their peers IOTA Desired Outcomes Recommenda[ons for source vendors Qualita[ve report that will help OpenURL providers Recommenda[ons for link resolver vendors? Beaer OpenURLs across the industry More success for users Current Work OpenURL repor[ng system: hap://openurlquality.niso.org/ Data (~ 9 million OpenURLs) from The Claremont Colleges Cornell University Georgia Ins[tute of Technology Kansas State University EBSCO Informa[on Services WorldCat Link Manager Serials Solu[ons Thomson Reuters Analy[cs / Reports Produce reports showing: Elements that are showing up in OpenURLs Paaerns of elements that are showing up in OpenURLs Frequency with which OpenURL elements (e.g., ar[cle [tle, ISSN, DOI, author last name, startpage, endpage, etc.) show up in OpenURLs from source databases and source vendors 3

4 Current Work Vendor completeness index: Tes[ng the assump[on that more informa[on makes for a beaer OpenURL Can we create a rela[ve ranking of source vendors based on how complete their OpenURLs are? Element weigh[ng: Which elements are present? Are some more important than others? How good are a par[cular vendor s OpenURLs? Issues Gathering and examining a lot of data Do the data accurately represent the user experience? Which OpenURLs (ejournals? ebooks? proceedings? everything?) Weigh[ng the elements Role of link resolvers/resolver selngs Case studies Upcoming Events Charleston Conference : Date: Friday, Nov. 2:00 PM Loca[on: Francis Marion, Parkview Room Speakers: Rafal Kasprowski - Rice University & Susan Marcin - Columbia University & hopefully other conferences soon. Stay tuned. 4

5 For more info Reports: hap://openurlquality.niso.org/ NISO page: hap:// Blog: hap://openurlquality.blogspot.com/ On Twiaer Follow IOTA on Twiaer: hap://twiaer.com/nisoiota To post comments on the IOTA page via Twiaer, use #nisoiota X KBART ok? X K? KNOWLEDGE BASES AND RELATED TOOLS: A NISO/UKSG RECOMMENDED PRACTICE Andreas Biedenbach KBART Working Group Co- chair October 13, 2010 NISO Webinar Agenda Who Publishers, aggregators, KB vendors, libraries What a universally acceptable holdings data format Where throughout the supply chain & at the UKSG info hub hap:// When Now Phase 1 Report Jan 2010 hap://bit.ly/kbartrp Phase 2 in progress Why Beaer access for users through accurate holdings data How can you help? For librarians For publishers Who is behind KBART? Standards organiza[ons UKSG and NISO The working group members, represen[ng: Knowledge Base vendors ExLibris, Serials Solu[ons, EBSCO Content aggregators Publishers Subscrip[on agents Libraries & consor[a Full list - - hap:// Phase II Working Group Jason Price Claremont Colleges / CDL Gary Pollack Cengage Learning/Gale Elizabeth Stevenson Edinburgh University Chad Hutchens University of Wyoming Sarah Pearson University of Birmingham Paul Moss OCLC Sheri Meares EBSCO Chris[ne Stohn Ex Libris Sherrard Ewing Serials Solu[ons Maahew Llewellin Royal Society Andreas Biedenbach Springer Marieke Heins Swets Ruth Wells Taylor & Francis Rose Robinson Publishing Technology 5

6 What is KBART? A simple metadata exchange format Knowledge Bases And Related Tools UKSG and NISO collabora[ve project UKSG 2007 research report: Link Resolvers and the Serials Supply Chain A set of prac[cal recommenda[ons for the [mely exchange of accurate metadata between content providers and knowledge base developers Representing years of thinking Where does KBART apply? date granularity (day, month, season, year) age of data file format format definitions; vol/issue vs date format frequenc y accuracy shoe-horning informal structure unclear accuracy link syntax and granularity incentive responsibilities Date coverage Data & transfer ownership duplication of effort title mapping contacts/feedback ISSN/ISBN variations Title relations Knowledge bases Supply chain mechanisms abbreviation s Licensing title changes Compliance accuracy re-use of format ISSN effect on licensing package free content variations genericism/ granularity misrepresentation What s in your knowledgebase the hard way Why KBART? Maintenance of accurate package content coverage data Supports OpenURL Link Resolvers Supports ejournal MARC record delivery services Enables automated upda[ng by KB providers Addresses common holding list inadequacies Re- use of ISSNs Embargo period ambigui[es Inconsistent date/enumera[on formats 6

7 How Librarians can help Lobby publishers to adopt the KBART prac[ces LEARN about what KBART is and what it does INSIST on knowing what you are buying! Require delivery of a usable holdings list Ini[ally & annually going forward When the list is inadequate, point them to KBART Only your insistence will make it happen ENABLE publisher sales staff to make the case to their company FOLLOW UP with con[nued requests as necessary Endorsers American Ins[tute of Physics, Ex Libris, OCLC, and Serials Solu[ons Announced by mid- September 2010: Alexander Street Press, Annual Reviews, EBSCO Informa[on Services, Innova[ve Interfaces, and Royal Society Publishing What to do to be endorsed? Providing KBART formaaed data Submit a sample file to the KBART working group at kbart@niso.org Endorsement is finalized once the file's format and content have been reviewed and approved and the provider has made it publicly available (in line with the recommenda[ons). What do publishers need to do to adopt the KBART best prac[ces? 1. Review the requirements that are accessible via hap:// 2. Format ejournal and ebook content availability data to meet the requirements. 3. Self check your datasheet(s) on the KBART website to ensure that they conform to the recommended prac[ce and make any necessary correc[ons. 4. Ensure that you have a process in place for regular data exchange as outlined in sec[on 5.2 of the KBART report. 5. Register your organiza[on on the KBART registry website (hap://bit.ly/kbartregistry) and provide a link to download the newly KBART formaaed dataset(s) The Registry - - a contact and metadata content clearinghouse Registry shortcut: hap://bit.ly/kbartregistry 7

8 Knowledgebase Management Before KBART Phase II Ac[on Items Proac[ve reconcilia[on of an ejournal package list Endorsement / Compliance / Engagement Defini[ons for global vs local updates Consor[a- specific metadata transfer Ins[tu[on- specific metadata transfer Review of metadata transfer for e- books Open access material General Process library, consor[um or KB vendor Request updated access list from publisher Sample publisher list for accuracy Translate publisher list to match KB list Number of [tles never matches Perform ISSN match with MS Access Watch for & integrate [tle changes, mergers, acquisi[ons and losses Watch for publisher- reuse of ISSNs/[tle combina[ons Iden[fy date discrepancies manually (inconsistent formats) Decide when it is good enough and go live/distribute new list Lather, Rinse, Repeat Our vision awer KBART Learn more hap:// Andreas Biedenbach KBART co- chair andreas.biedenbach@springer.com KBART interest group hap:// Phase 1 Universally accepted standardized publisher metadata, regularly distributed AND available on demand Phase 2 Broad adop[on, more content type coverage Phase 3? Consor[a & ins[tu[on level holdings metadata distribu[on based on what is actually accessible from a par[cular IP OpenURLs & Knowledge Base Quality: Collec[ng & Improving Knowledge Base Data Maria Stanton Director, Content Opera[ons October 13, 2010 SS_T.09 8

9 Agenda Patron Perspec9ve link leads to an ar9cle. " Patron Perspec[ve " Referring Source Data & the Knowledge Base " Target Source Data & the Knowledge Base " And the world would be a be/er place Patron perspec9ve When the link doesn t lead to the treasured resource they have no idea why all they know is, it didn t work 9

10 Trends Data Handshakes Alignment in 3 Spots Referring source Knowledge base Target source Data management and OpenURL " What does the referring source provide to support ar[cle level links? " What data does a typical target source provide to knowledge base providers? " Knowledge base provider our efforts to support the patron? OpenURL Standard ar9cle level Key parameters from the referring source Key rw.a[tle rw.[tle or rw.j[tle rw.date rw.volume rw.issue rw.spage rw.issn or rw.eissn rw.id Explana9on Ar[cle [tle Journal [tle (v 1.0 prefer j[tle) Date of publica[on Volume designa[on (number, roman numerals or non- numeric) Issue designa[on (number or non- numeric) Start page Standard iden[fier Iden[fiers unambiguously specific the referent ar[cle. Parameters = DOI (Digital Object iden[fiers) or PMID (PubMed Iden[fiers) What does a typical knowledge base provider track Full text coverage at the journal level Title Issn or eissn Full text start date Full text end date Journal- level URL Embargo informadon Knowledge base Sources of holdings data Electronic files: % are provided in a standard format (e.g., KBART) % are content provider specific Manually harvest: % of providers do not offer electronic files 10

11 Limita9ons of the files Most files don t include the full 9tle history EXAMPLE A system that doesn t iden9fy all three 9tles won t link correctly. Provider Title Freeman / , with coverage from June 1998 to present Actual Title Split Freeman / , coverage from June 1998 to Dec 1999 Other examples " Title Histories Provider reports following coverage: " R&D Magazine 1994 to the present " R&D Magazine was Research & Development un[l 2002 " Coverage dates Provider reports the following: " Public RelaDons Quarterly full text 1965 to present " Full text coverage ended in 2007 Ideas on Liberty / , coverage from Jan 2000 to Nov 2003 Freeman (2003) / , coverage from Dec 2003 to present Authority data management Titles & the accuracy of ISSN varies from provider to provider Title: Journal of the American Society of Nephrology Provider List Correc[ons Content J. Am. Soc. Nephrol Alternate Title JASN Alternate Title Results of the authority data management " Library s profile is accurate " Authorita[ve [tle and ISSN passed to target source Journal of the American Society of Nephrology Missing ISSN Journal of the American Society of Nephrology OK Journal of the American Society of Nephrology Error Journal of the American Society of Nephrology eissn Journal of the American Society of Nephrology : JASN Alternate Title Target source Target source s9ll resolve to correct item " Awer referring source and knowledge base hand shakes, need to call the target source " Knowledge base management focuses on journal [tle & coverage date issues " Link resolvers rely on IF/THEN scenarios " If a DOI is provided, use DOI to create ar[cle link " If DOI is not provided, but ISSN, VOLUME, ISSUE, and SPAGE are, use that metadata to provide an ar[cle level link. " Ar[cle [tle is typically problema[c 11

12 Ques9on Can we come together to do more? Helping Your Library Be The Best Partner For Research Thank You Contact us: Visit us: 12