Link Resolver Commercial Solution Selection, Implementation, Usability, Statistics, Development, Maintenance Frances Knudson Los Alamos National Laboratory.

Link Resolver Commercial Solution

Selection, Implementation, Usability, Statistics, Development, Maintenance

Frances KnudsonLos Alamos National Laboratory

Library Without WallsOctober 29, 2003LA-UR-03-7968

LANL Research Library

• LinkSeeker version 1.0 Nov. 2000

• 8 locally loaded databases

• Science Server – SDOS - 8 publishers

• Ejournals – 5,800 titles

• LANL electronic reports

Vocabulary• Source – starting point

• Target – what you use to get somewhere (ejournal, database, web)

• Object – journal – book – entity

• Services – actions

• Knowledge base – Stores connections between sources, targets, services, and objects.

Selection – Electronic Portfolio

• Databases – A&I, OPAC

• Ejournals

• Other types – books, reports, patents

• Ejournal listing products

• Proxy server

• Rank the sources and targets

Selection – Services

• What services do you want to provide? Full text linking, ILL (document delivery), OPAC look up

• Do you provide some of these services with other tools?

Selection – Level of Involvement

• How much customization do you want/need/desire?

• How much control?

• What is acceptable “turn around” time?

• Analyze computer skills of staff

Questions for Vendors

• Coverage of electronic portfolio

• Level of involvement

• Services

• Request current user names

• What are steps to add new sources, targets, objects?

• Update cycle for knowledge base

Implementation

• Assemble a team – ejournal/database, customer service, system staff, testers.

• Use the list of top ranked sources and targets. Determine cut off point for roll out.

• Determine what services are required for roll-out.

Oct. 2000 menu

Services we requiredfor roll out

We were offering full textlinks via another method

Implementation

• Start training staff

• Develop processes to move smoothly to normal work flow

• Allow user access in test mode

• Test, test, test

• Market, train

• Roll out

User interaction

• Focus groups – one on one & group sessions.

• First round was mid-2000. Every user suggested a special service. Links to Mathematical Reviews – links to their special web pages – BibTeX output - configurable menus.

• Continues on a daily basis – customer service staff visit staff, attend group meetings, gathering user feedback

Statistics

• During implementation phase – determine what you what to count – does resolver provide – if not – figure out how

• Usage – sources, targets, services

• Number of users

0

200

400

600

800

1000

1200

1400

1600

Jun-01 Sep-01 Dec-01 Mar-02 Jun-02 Sep-02 Dec-02 Mar-03 Jun-03

Date

Number of IPs

Statistics

• After 6 months – review

• Services – what are the most popular? What is their position in the menu?

• Sources, Targets – used – not used – good clues for training/marketing

• Users – is the number growing?

No document delivery service due to Full Text service

Access to JCR – new service

User requested service – download to EndNote

BioSciences Sources

0

100

200

300

400

500

600

700

800

900

1000

1100

Oct-01Nov-01 Dec-01 Jan-02 Feb-02 Mar-02 Apr-02 May-02 Jun-02 Jul-02 Aug-02 Sep-02

Oct-02Nov-02 Dec-02 Jan-03 Feb-03 Mar-03 Apr-03 May-03 Jun-03 Jul-03 Aug-03 Sep-03

BIOSIS PubMed

SciSearch Plus 14359

INSPEC 2014

Default 1421

PubMed 874

CAS 436

EngIndex 359

DOE Energy 295

Biosis 239

Science Server 144

NSA 122

Cambridge (CSA) 61

MathSciNet 16

getFullTxt 12589

getDocumentDelivery 787

getCitedReference 721

getCitedJournal 350

getCitedAuthor 186

getHolding 166

getReference 114

getAuthor 110

getWebSearch 108

getReport 86

getFullTxt2 38

getAuthorEmail 26

getLibrarian 19

getAbstract 17

Top sources and services for Sept. 2003

Development – Phase “n + 1”

• After initial rollout, rank remaining sources, remaining targets. Activate, test, move to production.

• Focus on refinements, new services, etc.

• Employ logic for services – if full text is available mask document delivery. Display preferred full text sources.

Development – Phase “n + 1”

• Review team

• Review assignments

• Refine processes created during implementation.

• Move to “normal” work.

Maintenance

• Monthly update from Ex Libris.

• Daily changes due to changes in collection

• Changes due to publisher changes

• Responding to users – links that do not work; do not work well enough, new services.

Maintenance

• Interconnected-ness (or interconnected-mess)

• If a link from the Link Resolver menu does not work, it is a problem of the Link Resolver.

• Browser problem, Acrobat problem, target site problem

• DOI – not registered

Developments at LANL

• Added OpenURLs to output from locally loaded databases. Services travel with the database record. Aug. 2002

• One lab group is using LinkSeeker in a local genome database.

• Mining LinkSeeker web logs for “recommendation systems”

Alert Email with OpenURL

• Article: http://linkseeker.lanl.gov/lanl?genre=article&issn=0253-6102&date=2003&volume=40&issue=3&spage=332&atitle=Structures+of+%28Omega+Omega%29%280%2B%29+and+%28Xi+Omega%29%281%2B%29+in+extended+chiral+SU%283%29+quark+model&aulast=Zhang&auinit=ZY Title: Structures of (Omega Omega)(0+) and (Xi Omega)(1+) in extended chiral SU(3) quark modelAuthor: Zhang, ZY; Yu, YW; Dai, LRJournal: COMMUNICATIONS IN THEORETICAL PHYSICS; SEP 2003; v.40, no.3, p.332-336

OpenURL included in all forms of output.

Due to feedback, a flag was added.

Default is to include OpenURL.

Need to accommodate intended use of output.

Enhancements

• Article level linking to Conference proceedings & books

• Archives – ever changing thresholds

• Ability to add local objects without impairing integrity of “Knowledge base”

• Statistics – “source to service” ratio – why?

Enhancements

• Databases – some are still not OpenURL aware

• Bibliographies – need more with OpenURLs

• Electronic books – OpenURLs in bibliographies

Frances Knudson

[email protected]

http://lib-www.lanl.gov