LIS 675 Database Content Evaluation Digital Fall 2012 Session 3. DATABASE CONTENT EVALUATION CRITERIA Book Chapter 2. Dr. Péter Jacsó Library and Information Science Program University of Hawai'i at M ā noa
Database Content Evaluation Criteria - Content through the prism of software The impact of interface and search engine Love/hate at first sight National anthem by Roseanne vs. W. Houston Sw intuitivity & user-friendliness help User guide can aggravate or alleviate problems Sw functionality can aggravate or alleviate problems JACSÓ
- Database Content Evaluation Criteria Browsable indexes make misspelled terms recognizable W/o browsing capabilities you are blindfolded At-once selection of variantly spelled terms eases the pain Truncation for one fell swoop retrieval of various endings No help if misspelling is at beginning or middle of term (Misubishi, Jasco) OCLC study result: most errors in 3 rd letter JACSÓ
Excessive misspellings of descriptors in the MHA database Jacsó 02-01
Basic index scanning in DIALOG online version of MHA - JACSÓ
- Basic index scanning in DIALOG online version of MHA JACSÓ
Selecting multiple index entries at a time in Information Science Abstracts no hit count Jacsó 02-02
How you select multiple entries at once in DIALOG? - JACSÓ
How you select multiple entries at once in DIALOG? - How you select multiple entries at once in DIALOG? JACSÓ
Excerpt from Patent Assignee Index in ISA disheartening Jacsó 02-03
Database Evaluation Process Casual use may give impressions Licensing requires systematic evaluation Battery of tests, experiments, analyses Value adding process also error adding process Lawsuit-proof contracts, no-one is responsible/liable Your contract reads as scripted POW statements for TV File producers and db publishers can get away with everything No malpractice suits, no false advertisement suits – yet Only exception D&B suit Fake PR statements about commitment to quality JACSÓ
Disclaimer from a Dun & Bradstreet database Jacsó 02-04
PR-spin JACSÓ
Counter spin Librarians, info professionals are the database industry final control team (Quint) Read some from these: Basch, Bates, Ojala, Pagell, Quint, Tenopir, and yours truly Would you spend your own money on this database? The power of publicity Not in vain, though slow process JACSÓ
The gratification JACSÓ
Hotlink for error notification from the Internet Movie Database Jacsó Hotlink for error notification from the Internet Movie Database
Database statistics in the Computer Science Bibliography database Jacsó 02-06
History and Literature of Content Evaluation Much to learn from print reviewers Surge in interest in warning users Anecdotic level, same old cliches Where have you been before the Web? Fear mongering against free resources Proportionally as many bad traditional commercial databases as free ones (not Web sites) Plain incompetence and ignorance JACSÓ
What did Jill discover? JACSÓ
A little memory reshuffling - JACSÓ
- A little memory reshuffling - JACSÓ
- A little memory reshuffling - JACSÓ
- A little memory reshuffling - JACSÓ
- A little memory reshuffling - JACSÓ
- A little memory reshuffling - JACSÓ
- A little memory reshuffling - JACSÓ
A little memory reshuffling
The classics - Martyn & Slater (1964): testing abstracting journals Lancaster (1971): Evaluation of indexes and abstract journals Bourne (1977): Spelling errors in databases Smith (1981): Incorrect citations distorting bibliometrics studies Williams & Lannom (1981): Lack of standardization in journal names JACSÓ
- The classics - Turtle & Robinson (1974): Time lag in LISA and LibLit Tenopir (1982): Methods of evaluating database coverage Gibson & Kunkel (1980): Covering Japanese sources by databases Ewbank (1982): Selecting databases Pemberton (1983): Dirty databases JACSÓ
- The classics - Comprehensive studies by Gilchrist, Goldstein, Dansey, Bottle & Efthimiadis, LaBorie, Halperin & White Drexel study about coverage and indexing in MEDLINE Sparck Jones (1976): Inadequacy of test collections Pao: Too small datasets JACSÓ
- The classics Jacso (1992): Census analysis of CD-ROM databases Hood & Wilson: LISA full scale descriptor analysis Traditional databases offer much better evaluation tools JACSÓ