The miracle of full-text searching: myth or reality?
Tim Hutchinson · 1999
Conclusions< Search interfaces/system design < Further research < Options for controlled vocabulary How do we measure retrieval performance?P Most established -recall and precision P Comparing systems/strategies -overlap rates P Other suggested techniques, e.g.failure analysis (qualitative), user-centred measures How do we measure retrieval performance?Problems/debates P "Real" users vs. subject specialists or information professionals P What is "relevance"?P Interactive searching vs. batch searching P Controlled conditions vs. "real world" Recall and precision Recall (# of relevant documents retrieved) -----------------------------