Information Retrieval LabExploring some IR toolsBasic InterfaceSome Features of EachExerciseCSC 9010: Special Topics, Natural Language Processing. Spring, 2005. Matuszek & Papalaskari1Information Retrieval LabCSC 9010: Special Topics. Natural Language Processing.Paula Matuszek, Mary-Angela PapalaskariSpring, 2005CSC 9010: Special Topics, Natural Language Processing. Spring, 2005. Matuszek & Papalaskari2Exploring some IR tools•You are probably all familiar with web-based IR tools, but we are going to play with three rather different ones today, mostly to look at their cababilities.–Google advanced search: http:www.google.com/advanced_search–Teoma: http://www.teoma.com/–MedLine: http://www.ncbi.nlm.nih.gov/entrez/query.fcgiCSC 9010: Special Topics, Natural Language Processing. Spring, 2005. Matuszek & Papalaskari3Basic Interface•Each of these tools has basic and advanced interfaces.–Try some queries on the basic interfaces –Then try some advanced query interfaces.•From Pubmed/Medline, click on the Limits tab to get to it.•What do they have in common?•How do they differ?CSC 9010: Special Topics, Natural Language Processing. Spring, 2005. Matuszek & Papalaskari4Some Features of Each•Google uses primarily a citation graph weighting to determine relevance, rather than keyword or other linguistic factors. •MedLine has a careful controlled vocabulary which can be used from the Limits tab•Teoma takes questions, not just keywords, and the results page includes a “Refine” section.CSC 9010: Special Topics, Natural Language Processing. Spring, 2005. Matuszek & Papalaskari5Exercise•Look for answers to these questions in the different engines: (some won’t give you much in MedLine)–What are effective treatments for Type 2 diabetes?–What towns are in North Carolina’s Outer Banks ?–Where can I open a checking account in Villanova?•And try some of your own.•What worked well in the different engines?•Did the special features
View Full Document