Clive Wright Senior Director of Discovery Products for Europe and Latin America
Web Scale Discovery: Abbr. Def. WSD systems couple… a pre-harvested central index of metadata and content… … with a richly featured discovery layer. 5 Sept Resource Discovery Services Athena Hoeppner University of Central Florida
Display Que? Catalog Metadata and Content Special Collections Search algorithm
Central Index Catalog Metadata and Content Special Collections Foundation Index Knowledge Base Pre-harvested Index
WSD/Vendor Agreements Publisher Supplied Library Subs Open Access FT and A/I Library Supplied Pre-harvested Index CENTRAL INDEX 5 Sept Resource Discovery Services | Athena Hoeppner
Fast, simple access to the library’s full-text content (electronic and print) in a single search box As well as… Your Library Catalog & Institutional Repositories Journals Conference Proceedings Magazines Newspapers Books Monographs Images Audio Video
Display Discovery Layer Search algorithm
How venture capital works. Author: Christian Science Monitor (1908- Current file) ISSN: Date: 03/11/1985 Description: 17 Search: Venture YU Library | Ask |Reserves | ILL | Your Account 700,000 hits Sort: Relevancy Dates Type Article Book Electronic …more Subject Finance Genetics Brothers …more Library Main A The Venture cloning method. Author: -Thadius S. Venture ISBN: Date: 2008 B Full Text Super Science Library QH 442 V Due: Yesterday!! Go Full Featured Discovery Layer 5 Sept 20138Resource Discovery Services | Athena Hoeppner
A discovery tool that is not discovering relevant content is worse than having no discovery tool But is it relevant? A discovery tool finding relevant content will improve results, speed up searches, use fewer clicks and increase usage of the most important content
Evaluating the Central Index Is the indexed content a good fit? – Inclusion of your journals titles and holdings – Inclusion of your subscribed databases Richness of metadata Ability to incorporate desired content in the future 5 Sept Resource Discovery Services Athena Hoeppner University of Central Florida
“If PsycINFO were included in discovery services, users could believe that they are actually searching the database, when, in truth, they would only be scratching the surface. We do not wish to contribute to such a misperception.” -Linda Beebe, Senior Director, PsycINFO Stakeholders Strive to Define Standards for Web-Scale Discovery Systems By Michael Kelley on October 11, 2012
“EDS is the only one that is integrated with a delivery (of PsycINFO)” -Linda Beebe, Senior Director, PsycINFO
EDS Record
Primo Central Record
Analyzing the Impact of Inclusion of a Major Subject Index NOTE: In Summon, the following expander was employed to increase results: “Add Results Beyond Your Library’s Collection” In Primo Central, the following expander was employed to increase results: “Include items not owned by the library” Searches conducted on April 25, 2013 NOTE: When retrieving zero results for adjacency searches, the Primo Central search engine sometimes ignores the quotation marks and returns a non-adjacency result list.
Display Discovery Layer Search algorithm
Evaluating the Discovery Layer Is it a good fit for your content? Is it a good fit for how your users search? Does it integrate with your other services? 20 5 Sept 2013 Resource Discovery Services Athena Hoeppner University of Central Florida
Relevancy and Value Ranking 21 5 Sept 2013 Resource Discovery Services Athena Hoeppner University of Central Florida Relevancy is extremely important and hard to pin down. The central index is a BIG pile of hay. The search and refinement tools are rakes The relevancy ranking is a magnet.
Relevancy and Value Ranking Weighted fields Exact match of term with field Density of terms Currency of item Document type Document length Locally owned 22 5 Sept 2013 Resource Discovery Services Athena Hoeppner University of Central Florida
Relevance Ranking EDS has the following priorities and has no bias toward content from any provider: 1) Match on subject headings from controlled vocabularies 2) Match on article titles 3) Match on author keywords 4) Match on keywords within abstracts 5) Match on keywords within full text Exact Matches: Exact matches are favored over partial matches – considering also the field in which those words appear (abstract vs. full text vs. title, etc.) Density: The number of times the word(s) appears relative to the size of the document (more is better) – considering also the field in which those words appear (abstract vs. full text vs. title, etc.)
Superior Value Ranking
Currency: When relevance is equal, the more recent journal articles rank higher Superior Value Ranking
Document Type: In relation to the search terms, document types may be relegated to lower relevance For example, when the word “book” or “review” is not searched, there is a “bias” against book reviews, so these don’t dominate results Superior Value Ranking
Length: Articles of a more substantial length have a heavier weighting (e.g., when relevance is equal, a ¼ page article is considered less valuable than a 4 page article) Superior Value Ranking
Usage of library content Goes Up with Discovery, but… …which content types experience an increase in usage depends on which discovery service is exposed to your end users
University of North Florida (USA) reported an increase in ScienceDirect usage of 247% due to the implementation of EDS (previous year only 11%) Case Study: EDS Increases Usage of a Top Publisher’s Journals Bournemouth University (UK) reported an increase in ScienceDirect link-outs of 547% over the previous year without EDS
Picking a WSD and Making It Yours 1.Know what you want to do with the service. 2.Pick the central index that aligns with your content and has the metadata to support discovery. Turn data sources on or off to focus the index for your users. 3. Pick a discovery layer that can surface the content your users want to find and that facilitates search behavior you want to promote Sept 2013 Resource Discovery Services Athena Hoeppner University of Central Florida
Muchas Gracias Clive Wright