federated search of today and tomorrow shanyun zhang electronic resources librarian, john k. mullen...
TRANSCRIPT
Federated Searchof
Today and Tomorrow
Shanyun Zhang Electronic Resources Librarian, John K. Mullen Library
The Catholic University of America
Potomac Valley Chapter of ASIST September 27, 2006
Outline
Summary of Our Research Project Findings Today’s Federated Search Possible Directions
Context
User interest in one-stop shopping Increased awareness among librarians of
the value of federated search WRLC implementation
Research Objectives
To understand users’ perceptions of a federated search system
To compare and contrast the perceptions of librarians and students regarding federated search
To identify areas of confusion and misperceptions and explore their implications for literacy programs in federated search and for system design
Methodology (1)
Survey instrument Experience with MetaLib Combined Search
(MCS) Usage Opinions of MCS Search simulation (see sample screens)
Narrative Drawing
Screen A
Methodology (2)
Participants Librarians from WRLC libraries Students from the School of Library &
Information Science MetaLib expert implementer for WRLC
Methodology (3) Data collection
Mail survey (early Jan. 2006) Take-home survey (late Jan. 2006)
Data processing & analysis RAs Descriptive statistics Content analysis of open-ended questions Analysis of components and relationships in
narratives and drawings (in progress)
Participants Background Total 19 Librarians
Total 22 students
Findings: Usage
32%
32%
0%
39%
23%
85%
0% 20% 40% 60% 80% 100%
Quick search
After fruitlesssearch
Full text
Librarians Students
Findings: Assessment
Useful, but complex and hard to figure out
Students: 69% Librarians: 11%
Not useful, and hard to figure out
Students: 8% Librarians: 47%
42% of the librarians believed users made use of MCS without knowing exactly how it worked.
0.00%10.00%20.00%30.00%40.00%50.00%60.00%70.00%80.00%90.00%
Students
Librarians
Findings: FindIt Function
Only 27% of the students noticed the “FindIt” button
Findings: Areas of Confusion(selected subcategories)
Librarians Students
Background informationHow MCS works and its limitationsDatabase information
44%26%
13%
42%12%
21%
MetaLib system operationResult display, ranking & relevanceFindItFull text retrieval
51%21%
18%
8%
27%15%
3%
9%
Search supportSearch tutorial, Search modification, Technical support, search history
5% 27%
Topics to Learn(selected subcategories)
Answers from Students
Background informationWhat databases are covered by MetaLibWhat full text journals users have access to Federated search process & what are differences
50%28%
9%
9%
MetaLib system operationhow return search results, how determine relevance
22%
Search strategiesHow to enter search terms; How to use search fields; How to broaden the search; How to use controlled vocabulary
13%
Practice 16%
Topics to Cover
Students want to learn: Background (50%) MetaLib system
operation (22%) Search strategies
(13%) Practice (16%)
Librarians want to teach: Background (30%) MetaLib system
operation (28%) Search strategies (37%) Instruction methods (5%)
Today’s Federated Search
Students say “Cool!” One interface One search to full text
Librarians are unhappy with its performance
Database inclusion Search speed Retrieval precision Result comprehensiveness
Possible Directions
System architectures Distributed search model (queries to multiple remote
resources in real time, which suffers from multiple limitations: number of connections available, slowness of remote services, and large result sets)
Centralized search model (gathering data on the universe of interest in advance and processing it into indexes that can provide instant results to searchers’ queries)
Cooperation between giant aggregators and publishers CSA MultiSearch
Information Standards
Z39.50 is the primary standard utilized in Metalib Difficult to sort large result sets Many vendors don’t support Z39.50
NISO Metasearch Initiative NISO Metasearch XML Gateway
Simplified CQL (Common Query Language) and SRU (Search & Retrieval via URL) for queries
XML for responses http://www.niso.org/committees/MS_initiative.html
Which is better?
Advantages DisadvantagesCommercial Federated Search Systems
Multiple search options Expensive; More maintenance staff needed
Google Scholar Free; Will be benefit from the open contents
Keyword search only; Cooperation with libraries needed
Cooperated Aggregators and Publishers
Metadata and content available;
Minimum maintenance in libraries
Expensive
Thank You!
Contact Information
Shanyun [email protected]
Coming publication: Tang, Rong, Hsieh-Yee, Ingrid, & Zhang, Shanyun. (Accepted). “User Perceptions of MetaLib Combined
Search: An Investigation of How Users Make Sense of Federated Searching”. Internet Reference Services Quarterly.
Future Readings
Breeding, M. 2005. Plotting a new course for metasearch. Computers in Libraries 25(2): 27-29.
Webster, P. 2004. Metasearching in an academic environment. Online 28(2): 20-23.
Sadeh, T. 2006. Google Scholar versus metasearch systems. HEP Librarians Webzone 12. http://library.cern.ch/HEPLW/12/papers/1/.
Cervone, F. 2005. What we’ve learned from doing usability testing on OpenURL resolvers and federated search engines. Computers in Libraries 25: 10-15.