Computers in Libraries Report
From PASCAL Wiki
"Super Searcher Shares 30 Search Tips," Mary Ellen Bates
CustomizeGoogle.com Firefox fix for Google: removes ads, infinite scroll on results
Google new experimental search add view:timeline or view:info to search query; shows images on page
Searchmash.com Results sorted by web page, images, blogs, videos, Wikipedia. Tool gives patrons more advanced search without lots of explanation.
Gigablast Clustering, "extremely cool complex search and ranking tools."
Exalead Forgiving with phonetic spelling, searches forbidden terms, supports old-fashioned boolean logic.
Maps.ask.com Ask is rebranding itself to a limited demographic ("women's search engine") but offers walking directions that take local topography into account.
SnapSearch Based on Gigablast, has their advanced features. Allows quick page preview without going to all pages.
NationMaster.com/ Freakonomics for librarians: heat maps, statistics, individual states.
Kosmix.com Assimilates information from a variety of sources, excellent clustering; focus currently limited to health, travel, and autos.
"What's New with Federated Search?" Presenters: Frank Cervone, was at Northwestern University now at Chicago State U; and Jeff Wisniewski, U Pittsburgh.
University of Pittsburgh Digital Library is in the process of implementing a new interface in next 2 months, will be WebFeat but with a new skin. No where on home page do they offer an A-Z list of databased. Cited College and Research Libraries article from November 2007 which found little difference between the quality of FS results when compared to native searches; in the study faculty rated the quality of FS results as good; faculty rated results higher than librarians. One wonders if this is because librarians know that other resources might be more appropriate for the topic.
Libraryfind at Oregon State. Built on open URL resolver; 2-click find workflow; developing ability to incorporate local indexes.
Simon Fraser University DbWiz is oldest of all, used in Canada.
Care Affiliates unbundling interface and connectors from back end database.
Masterkey LibLime product, almost instant search of up to 100 databases, most recently developed OS Federated Search product. Returns 2000 records per second.
Next generation OPACs with FS: vufind, WorldCat Local, Encore, Primo, Aquabrowser. Chicago State University's experience with Aquabrowser is that faculty and graduate students are finding things they wouldn't find otherwise because it maps to non-traditional relationships which aren't reflected in catalogers' thinking.
Federatedsearchblog.com Vendor-neutral, searchable, comprehensive.
Trends: moving toward a more holistic approach, data pre-processing approaches, software as a service with remote hosting. Development of standards, SRU & SRW Federatedsearchblog has an explanation.
“User Generated Content,” Roy Tennant
Tennant wrote forward to Social Software in Libraries. The book discusses user-generated content on library-managed sites, distinguishing between “real” user-generated content and descriptive content. Real content being photos, documents and the like; descriptive content are tags, ratings, etc.
Why bother with user generated content? More (decent) content is better, more access is better. It fosters interaction and community, provides more personalized service. "More data trumps better algorithms."
Contributions and content examples: eScholarship Repository at University of California system, where faculty (well, administrative assistants and graduate students, really) are actively depositing papers into archive. Kete project in New Zealand, where communities are contributing to a genealogical database, identifying people & places in historical photograph collections, contributing to encyclopedias, and have a model real estate site, and much more.
Another example: when Library of Congress put images on Flickr, user comments helped with corrections, adding names of previously unidentified people, making real contributions to the metadata. Using Flickr, Library of Congress got more than 10,000 unique tags out of 55,000 total, they got more than 11,000 friends/contacts, and more than 3,500 comments were posted by more than 1,400 users. Tennant's point was that Flickr deepened LC's content but also opened LC's virtual doors to vast numbers of users.
Also WorldCat “Add to it” review and public note at Boston Public Library.
Exploiting the knowledge of the masses enriches content, distributes value-added possibilities beyond library staff.
Contributions to Discovery
Bookspace Hennepen County Library with WorldCat allows users to make lists and add tags. User terminology is “low barrier,” low overhead, but tags don’t deal with pharases well.
LibGuides and LibMarks. LibGuides for librarians, LibMarks is social bookmarking for libraries. With LibraryThing it allows you to meld your opac with new content without buying a next gen catalog, i.e., “put the pig in a dress.” (Roy was embarrassed to have said that but it was still amusing.)
“Get used to being in a messy world; our idea of content may not be theirs.” User engagement is good, it gets people thinking about libraries. Library OPACs look like fortified castles to users now.
On consortial tagging, Roy says the smaller you make your pool, the less effective it will be, so he recommends sharing tags.
"Next Generation Library Interfaces," Marshall Breeding
OPACS are not good at delivering e-content; it's time to raise the bar, decouple the front end and the back end.
- more comprehensive information discovery environments
- digital resources can't be an afterthought
- e-content only is also a mistake
- forcing users to use different interfaces depending on content type is becoming less tenable.
- libraries are working toward a consolidated user interface.
Ideal is to move not toward federated search but more like Open Archives Initiative. Our primary automation systems should have web 2.0 features built in from the beginning, not as an afterthought, creating information silos.
Characteristics of ideal next generation software:
- unified user experience with a single entry point to all content and services with optional advanced search
- print and electronic
- local and remote
- locally created content
- user-contributed content
- relevancy ranking: From web, users expect relevancy ranking; Endeca & Lucene cited as models. Good RR requires sophistication and should factor in objective matching criteria but also popularity and relatedness factors; social factors should be considered.
- single sign-on
- facets for narrowing and navigation
- query enhancement such as spell check and "did you mean." We should automatically provide authorized terminology and related terms; we have an abundance of metadata, so we should make the query and response to the query more useful than the original query.
- breadcrumbs
- visual content
The web implies a new paradigm of web searching, letting users drill down incrementally narrowing; advanced boolean and keyword are dying, so we should give tools that acknowledge that the user isn't using boolean and keyword. Advanced boolean requires that the user already know what you have in your opac. Drill-down method gives users number of hits in each facet or topic. Facets mimic online shopping experience, support the ability to explore without a priori knowledge.
Deep search: we're entering the post-metadata search era, with full content initiatives where we'll be able to search digital object whether print, sight, or sound. It's embarrassing that Google can offer search inside a book and we can't.
Only full-text searching without metadata is also a loss; we need both.
Example of appropriate relevance factors: objective keyword rank + library weight + circulation frequency + OCLC holdings + scholarly content.
Offer groupings that are less "clutterful," FRBR.
Aim for enterprise integration: courseware, social networking, campus portal, feed aggregators ... not just a dumbed down OPAC.
"One year on the web is like seven years in library time."
"Demystifying and Integrating Web 2.0 at the Metropolitan Museum of Art," Deborah Vincelli and Jennie Pu
Started with a general call to all Met library staff to see who was interested in exploring 2.0 tools to see if they would be useful for curators and art historians. Selected 7 tools, used the tools to explain them in a guide, making a page for each tool. Avoided industry terminology, showing blogs created by people in similar jobs.
The committee used Wordpress to track committee notes, this cut down on physical meetings; they created guidelines for sharing information and determining what information goes where.
They use Google calendar and Google documents for schedules; Google Notebook to share notes.
del.icio.us - demonstrated that scholars use it to share resources and prints
Flickr - recommended for sharing fieldwork; they found educational lectures, use by other museums in an official capacity - Brooklyn Museum.
RSS - scholarly art history journals using RSS feeds to offer abstracts and full text of articles long before actual print publications are released.
Had to convince scholars that "content can be public, completely private, or restricted to a select group of colleagues, you decide."
Case Western wiki can only be edited by individuals.
Previously unknown scholarly content found by art historians using the bibliography of Wikipedia.
They use Google chat and Meebo for internal communication among Met library staff.
Art Library Society of North America using Flickr, RSS, blog
Art historians and curators shifted from an attitude of suspicion about 2.0 tools to being active contributors to 2.0 bodies of knowledge.
"Innovative Marketing Using 2.0 Tools," Helen Blowers and Mike Porter ("Libraryman")
The old marketing paradigm: control the look and feel of our brand. The new paradigm: influence the character and community of the brand. Encourage portability of the brand, enable participation with the brand, encourage consumers to participate in the creation of the brand.
You prove your high value when your users adopt your brand.
Allen County Public Library - won an InfoTubey in 2007 for their community album, "A day in the Life of Allen County."
Make a Wikipedia page for your library.
If you put anything on youtube make it searchable.
City of PaloAlto's wiki for book reviews, blogs.
Hennepin County Public Library recommended text for granting rights to use photos of users.
Three reasons you love your library at NJ State Library: hugely successful, innovative, youtube project.
Big Huge Labs - image generator allows you to create your own "motivational poster"
Denver Public Library YouTube contest "How I had Fun in the Library" for teens, video had to be inside or outside the library and couldn't be about sex -- only rules.
Eight steps to marketing with 2.0 tools: 1. learn social media, educate yourself. 2. use it. 3. envision how to apply it, shift the focus on marketing to participation and conversation 4. engage, create social celebrations focused on the customer 5. enable portability of the brand 6. play with multimedia 7. explore as you go and track success 8. experiment
Question about use in academic libraries, Western Washington University library is the main thing that shows up on youtube search because of their use of it in library.
Google Tracking, Greg Notess from Search Engine Showdown
User interface tests at Google are constant and vary by user, users IP affiliation. Search results will vary with user's profile.
Search in site feature: Library of Congress and New York Times - results aren't as good as native search.
Recommends using Ask for images, maps.
Exalead is the advanced searcher's favorite, but the database is still out of date.
What's Hot according to Steven M. Cohen
http://www.slideshare.net/ Slideshare
