Document Actions
CIL2008--Next-Generation Library Interfaces
April 08, 2008 03:31 PM | CIL2008Donovan Lambright, Automation Librarian
NOTE: The WiFi in the conference rooms has been absolutely terrible today and I am seriously behind in getting my session notes posted. I'll get caught up but it will take a little extra time. Computers in Libraries really needs to address this next year.
Next-Generation Library Interfaces
Marshall Breeding, Director for Innovative Technologies and Research, Vanderbilt University
From the program:
"Following a period of widespread dissatisfaction with library OPACs, a number of projects and products have emerged to provide library interfaces more capable of satisfying the needs of web-savvy library users. New models of what constitutes a library catalog are forming, and products are now beginning to embody aspects of this new vision. Breeding provides an overview of the library catalogs and interfaces now available or in development, including both commercial and open source alternatives."
In 2005, OCLC found that 2% of students think to go to the library catalog first when they need information. 89%, on the other hand, go to search engines. The numbers for non-students were worse with 1% going to the catalog first. Since then, followup surveys have shown that our numbers have gone even further down.
In short, we have competition in the web and the web is winning.
The problem and moves toward a solution
Libraries need:
The inadequacies of ILS OPACs have been well defined in the last few years by many writers and speakers:
Change is underway in the library community:
The next-generation of OPACs:
How do we create next-generation OPACs?
Redefine the OPAC:
Create a comprehensive search service:
Include Web 2.0 functions:
New technology will be necessary to support new functionality:
What should these OPACs look like?
The ideal scope of next generation OPAC interface:
Functions and features
Interface should include:
Relavancy ranking:
This requires a new paradigm for search and navigation:
The next-generation OPAC must enhance the patron's queries so they don't have to:
Behind the scenes, the appropriate organizational structures must be in place:
Enriched content is vital:
Patrons used to the web expect a personalized user experience and single sign on:
Deep search will become increasingly important:
Next generation OPACs must look beyond discovery
We can't just imitate Amazon and Google; we need library-specific features:
Must integrate in the enterprise:
Interoporability:
Other considerations
Great benefit, great costs:
Can we afford a slow transition?
Open source opportunities:
Current solutions:
For more information, see June/July 2007 issue of Library Technology Reports.
NOTE: The WiFi in the conference rooms has been absolutely terrible today and I am seriously behind in getting my session notes posted. I'll get caught up but it will take a little extra time. Computers in Libraries really needs to address this next year.
Next-Generation Library Interfaces
Marshall Breeding, Director for Innovative Technologies and Research, Vanderbilt University
From the program:
"Following a period of widespread dissatisfaction with library OPACs, a number of projects and products have emerged to provide library interfaces more capable of satisfying the needs of web-savvy library users. New models of what constitutes a library catalog are forming, and products are now beginning to embody aspects of this new vision. Breeding provides an overview of the library catalogs and interfaces now available or in development, including both commercial and open source alternatives."
In 2005, OCLC found that 2% of students think to go to the library catalog first when they need information. 89%, on the other hand, go to search engines. The numbers for non-students were worse with 1% going to the catalog first. Since then, followup surveys have shown that our numbers have gone even further down.
In short, we have competition in the web and the web is winning.
The problem and moves toward a solution
Libraries need:
- interfaces our patrons like to use
- powerful capabilities in tune with how the web works
- interfaces that meet expectations that are usually set by the web
The inadequacies of ILS OPACs have been well defined in the last few years by many writers and speakers:
- not great at delivering electronic content
- based on complex text-based interfaces
- weak keyword searching
- lack of good relevancy sorting for search results
Change is underway in the library community:
- widespread dissatisfaction with the current generation of OPACs
- movement among librarians to decouple the OPAC (front end) with the ILS (back end)
- vendors moving (slowly) to redesign the ILS to be more effective with both digital and print content
The next-generation of OPACs:
- redefine the library catalog and question traditional notions of searching
- provide better information delivery tools
- offer more powerful search capabilities
- offer more elegant interfaces
How do we create next-generation OPACs?
Redefine the OPAC:
- provide more comprehensive information discovery environments
- it is no longer enough to limit to print resources; digital resources cannot be an afterthought. On the other hand, books are not going away either. A balanced approach is necessary.
- forcing users to use different interfaces depending on type of content (books, online databases, etc) is no longer tenable
Create a comprehensive search service:
- not federated search (which is a step in the right direction but not the ultimate answer) but more like OAI (Open Archives Initiative) with consolidated search based on metadata and data gathered in advance
Include Web 2.0 functions:
- more social and collaborative approach
- web tools built right in:
- blogs, wikis, tags, social bookmarking, user ratings and reviews are all integrated, not an afterthought
- should avoid web 2.0 silos--integrate the user-created content, don't just put in in a sidebar like an add-on
New technology will be necessary to support new functionality:
- web services
- xml apis
What should these OPACs look like?
The ideal scope of next generation OPAC interface:
- unified user experience with single point of entry
- works well with print and electronic resources
- search and integrate local and remote resources
- search for library-created content
- integrate user-generated content
Functions and features
Interface should include:
- simple point of entry for searching with advanced search optional
- results ranked by relevancy
- faceted search for narrowing and navigation
- query enhancement--spell check,for example
- suggested results related to initial query for browsing
- navigation bread crumbs
- enriched visual and text content
- syndetics
- amazon
- google book search
- librarything for libraries
- single sign on regardless of where patron "moves"
Relavancy ranking:
- based on advanced search engines designed with relevancy ranking as a key function such as:
- endeca
- lucene
- web users expect this as a matter of course
- the good stuff should be listed first
- users won't delve into a deep list
- requires a sophisticated approach, with objective matching criteria supplemented by popularity and relatedness factors
- e-commerce solutions have simple goal--what they want most to sell goes at the top of the list. Libraries have a different set of values and our goal, getting patrons the most appropriate information, is a lot tougher. That doesn't mean it can't be done or that we should stop trying.
This requires a new paradigm for search and navigation:
- patrons tend to start with broad, general searches and drill down to find what they want
- librarians tend to start with narrow, precise searches so that no further refinement is necessary after the results are found
- we aren't going to persuade users to search like we do. more bibliographic instruction is not the answer.
- faceted browsing, visual search, and navigational breadcrumbs are all tools that will help
The next-generation OPAC must enhance the patron's queries so they don't have to:
- "did you mean?" and other features to help avoid the dreaded "no results found" message
- validated spell check
- automatic inclusion of authorized and related terms
- "more like this"--recommendations of similar items
- make the query and the response to it better than the query provided by the patron
Behind the scenes, the appropriate organizational structures must be in place:
- Library of Congres Subject Headings vs fast faceted application of subject terminology
- full MARC vs Dublin Core or MODS
- discipline-specific ontologies such as BISAC, the subject headings used by bookstores
- tags
- we must get away from the "not invented here syndrome". MARC and LC are not the sole repositories of that which is good.
Enriched content is vital:
- rich visual information such as book jackets, reviews, etc.
- syndetics (costs money)
- amazon web services (free but comes with Terms of Use requirements that seems to preclude libraries from using it)
- google book search api (free but just released in march 2008, so the jury is still out on its usefulness; initial impressions are good)
Patrons used to the web expect a personalized user experience and single sign on:
- customized content options based on personal preference and profile of previous use
- persistent sign on
- seamless navigation through subsystems
- credentials that follow user regardless of service or materials being searched
- ability to select and save content for later use
Deep search will become increasingly important:
- we are entering post-metadata search era with increasing opportunities to search full content
- google liberary print
- google publsher
- open content alliance
- microsoft live book search
- commercial websites like amazon already offers "search inside the book". Why can't we?
- although it is important, deep search is not the whole answer; it is highly improved by good metadata. we still need catalogers.
Next generation OPACs must look beyond discovery
- fulfillment orientated; finding the item is just the first step--we have to put in the patron's hand and/or monitor
- search-select-view
- delivery/fulfillment is harder than discovery
- back end complexity associated with delivery should be seamless for patron
- offer services that work equally well for digital and print content
We can't just imitate Amazon and Google; we need library-specific features:
- appropriate relevance factors for a library are not the same as those for a commercial enterprise
- keyword ranking plus library weightings
- circulation frequency
- OCLC holdings
- scholarly reputation
- results grouping (FRBR)
- we are collection focused, not sales-driven. we don't have to try to be google
Must integrate in the enterprise:
- ability to deliver content and services through non-library applications
- campus portal solutions
- courseware
- social networking sites
- search portals
- rss feed aggregators
Interoporability:
- decoupled OPAC (not part of the ILS) implies data synchronization
- regular mass export catalog data to the OPAC for indexing
Other considerations
- next-generation OPACSs must be smart and sophisticated
- creating them is much more difficult than old generation OPACs
- they are not a "dumbed down solution" as some more traditionally-minded librarians have described them; quite the contrary
- represents a merger of library requirements with new patron expectations with e-commerce solutions
Great benefit, great costs:
- represent a whole new level of expense
- costs of ils OPACs are modest compared to next-generation OPACs
- can libraries bear the cost?
- can we afford not to?
Can we afford a slow transition?
- libraries tend to move slow
- the web tends to move fast
Open source opportunities:
- not that many deployments yet
- time to market is critical
- challenge to catch up quickly
Current solutions:
- endeca
- encore
- primo
- lucene
- aquabrowser
- oclc worldcat local
- indigo
- librarything for libraries
- scriblio
- vufind
- extensible catalog
- polaris
- koha
- evergreen
For more information, see June/July 2007 issue of Library Technology Reports.
Comments
There are no comments yet.
