Presentation is loading. Please wait.

Presentation is loading. Please wait.

© Tefko Saracevic, Rutgers University 1 EVALUATION in searching IR systems Digital libraries Reference sources Web sources.

Similar presentations


Presentation on theme: "© Tefko Saracevic, Rutgers University 1 EVALUATION in searching IR systems Digital libraries Reference sources Web sources."— Presentation transcript:

1 © Tefko Saracevic, Rutgers University 1 EVALUATION in searching IR systems Digital libraries Reference sources Web sources

2 © Tefko Saracevic, Rutgers University 2 Importance of evaluation  Integral part of searching  always there - wanted or not  informal or formal  Growing problem for all  information explosion makes finding “good” stuff very difficult  Formal evaluation part of professional job & skills  requires knowledge of evaluation criteria, measures, methods  more & more prized

3 © Tefko Saracevic, Rutgers University 3 Information systems  Considered here:  information retrieval (IR) systems, e.g. Dialog, Nexis …  sources included in digital libraries, e.g. Rutgers  reference services e.g. in libraries or commercial on the Web  Web sources e.g. as found in many domain sites  Many approaches, criteria, measures, methods are similar & can be adapted for specific source or information system

4 © Tefko Saracevic, Rutgers University 4 Broad context Evaluating the role that an information system plays as related to: èSOCIETY - community, culture, discipline... èINSTITUTION - university, organization, company... èINDIVIDUALS - users & potential users (nonusers) Roles lead to broad, but hard questions as to what context to choose for evaluation

5 © Tefko Saracevic, Rutgers University 5 Context (cont.)  Social:  how well does an information system support social demands & roles? hardest to evaluate  Institutional:  how well does it support institutional/organizational mission & objectives? tied to objectives of institution also hard to evaluate  Individual:  how well does it support inf. needs & activities of people? most evaluations in this context

6 © Tefko Saracevic, Rutgers University 6 Approaches to evaluation  Many approaches exist  quantitative, qualitative …  effectiveness, efficiency...  each has strong & weak points  Systems approach prevalent  Effectiveness: How well does a system perform that for which it was designed?  Evaluation related to objective(s)  Requires choices: Which objective, function to evaluate?

7 © Tefko Saracevic, Rutgers University 7 Approaches … (cont.)  Economics approach:  Efficiency: at what costs?  Cost-effectiveness: cost for a given level of effectiveness  Ethnographic approach  practices, effects within an organization, community  learning & using practices & comparisons

8 © Tefko Saracevic, Rutgers University 8 Basic requirements for evaluation Once a context is selected need to specify all five: 1. Construct  A system, process, source e.g. a given IR function or system; a Web site, a Dlib source 2. Criteria - to reflect objective(s)  e.g. relevance, utility, satisfaction, accuracy, completeness, time, costs 3. Measure(s) - to reflect criteria  precision, recall, various Likert scales, $$$,...

9 © Tefko Saracevic, Rutgers University 9 Requirements … (cont.) 4. Measuring instrument - judgments by users on relevance or on a scale; cost/function 5. Methodology - procedures for collecting & analyzing data  No evaluation can proceed if not ALL of these are specified!  Sometimes specification on some are informal & implied, but they are always there.

10 © Tefko Saracevic, Rutgers University 10 In IR: 1. Construct  In research: most done on test collections & test questions  Text Retrieval Conference - TREC evaluation of algorithms, interactions reported in research literature  In practice: on use & user level: mostly done on operational collections & systems  e.g. Dialog, Nexis, various files evaluation, comparison of various procedures, commands, contents user proficiencies, characteristics evaluation of interactions reported in professional literature

11 © Tefko Saracevic, Rutgers University 11 2. Criteria  Relevance basic & most used criterion  strengths, weaknesses  Relevance as area of study  basic notion in inf. science  User & use level: many other  utility, satisfaction, success, time, value, impact...  Market evaluations:  those + quality, fitness-of-use, penetration...

12 © Tefko Saracevic, Rutgers University 12 3. Measures  Precision & recall preferred  Problem with recall: how?  use of methodological “tricks”  some think as metaphysical  Use & user level  Likert scales, differentials for many criteria e.g. satisfaction on a scale of 1 to x (1=not satisfied, x=satisfied)  observational measures e.g. overlap, consistency

13 © Tefko Saracevic, Rutgers University 13 4.Instruments  People used as instruments  they judge relevance, scale...  But people who?  users, surrogates, analysts, domain experts, librarians...  How do relevance, utility... judges effect results?  who knows?  Reliability of judgments:  about 50 - 60% for experts

14 © Tefko Saracevic, Rutgers University 14 5. Methods  Includes design, procedures for observations, experiments, analysis of results  Challenges:  Validity? Reliability? Reality? Collection - selection? size? Request - generation? Searching - conduct? Results - obtaining? judging? feedback? Analysis - conduct? tools? Interpretation - warranted? generalizable?

15 © Tefko Saracevic, Rutgers University 15 Criteria for evaluation of information sources  Includes Dlib & Web sources  What?  Content: Subject? Topic?  Level? Depth? Exhaustively? Specificity? Organization?  Timeliness of content? Up-to- date? Revisions? Accuracy?  Why?  Purpose? Scope? Viewpoint?  For?  Intended audience? What need satisfied? Appropriateness?

16 © Tefko Saracevic, Rutgers University 16 Criteria...  Who done it?  Author(s), institution, company, publisher, creator: Authority? Reputation? Credibility? Persistency? Trustworthiness? Refereeing? Transparency?  How?  Content treatment: Readability? Style? Organization? Clarity?  Physical treatment: Format? Layout? Legibility? Visualization?  Where?  Availability? Accessibility?

17 © Tefko Saracevic, Rutgers University 17 Criteria...  How?  Searching, navigation, browsing?  Feedback? Links?  Output: Organization? Features? Variations? Control?  Effort? Learning factors?  How much?  Price? Total costs? Cost- benefits?  In comparison to?  Other similar sources?

18 © Tefko Saracevic, Rutgers University 18 Conclusions  Evaluation is a complex task  but also an essential part of being an information professional  Traditional approaches & criteria still apply  but new ones added or adapted to satisfy new sources, & new methods of access & use  Evaluation skills in growing demand particularly because Web is value neutral  Great professional skill to sell!


Download ppt "© Tefko Saracevic, Rutgers University 1 EVALUATION in searching IR systems Digital libraries Reference sources Web sources."

Similar presentations


Ads by Google