Key Word in Context Concordances
from the Web with

For historical purposes only—
KWiCFinder is no longer maintained or functional

• Sample search reports in French, German, Spanish and Dutch
• Publications / presentations
  • Implementing a BNC-Compare-able Web Corpus" (PDF of paper), Web as Corpus 3, Louvain-la-Neuve BE, 15-16 September 2007.   Presentation (PDF of PowerPoint)  photos
  • "Toward cleaner Web corpora: recognizing and repairing problems with hybrid online documents", (PDF of PowerPoint)  Corpus Linguistics 2007, Birmingham UK, 27-30 July 2007. photos & videos
  • Concordancing the Web
    "Concordancing the Web: Promise and Problems, Tools and Techniques"   (Acrobat .PDF)
  • Identifying and Researching Multi-Word Units, BAAL and OTA, Oxford 21 April 2005
    "Real-Time Identification of Multi-Word Expression Candidates in Databases from the British National Corpus and the Web"   (Acrobat .PDF)
  • Web as Corpus SSMILT Forlì 2005
    "Towards an Independent Search Engine for Linguists:  Issues and Solutions" Acrobat .PDF
  • 2002 North American Symposium on Corpus Linguistics    PowerPoint   
    "Making the Web More Useful as a Source for Linguistic Corpora" (Revised paper .PDF) n-gram data (.PDF, 2 MB)
  • Teaching and Language Corpora 2002    PowerPoint    TaLC 2002 Photos
    "Facilitating Compilation and Dissemination of Ad-Hoc Web Corpora" (Revised paper .PDF)
  • Applied Corpus Linguistics 2001    PowerPoint | HTML
    "Concordancing the Web with KWiCFinder" (Draft paper .PDF)
  • CALICO 2001   Acrobat .PDF | HTML 
  • EuroCALL 2000   PowerPoint | HTML
  • Didascalia-Symposium 1999 (in Dutch)   PowerPoint | HTML

The Solution...

• KWiCFinder conducts your online searches without supervision
• It returns Key Word in Context abstracts highlighting your search terms so you can evaluate the usefulness of documents matching your query
• Search reports are highly interactive [details]
• Numerous enhancements permit more finely-targeted searches than typical web search engines

The Problems...

• So many hits... How can you tell which are relevant?

  Eliminate most irrelevant hits with highly-specific searches, then reject unwanted matches (retrieved and analyzed automatically) based on your assessment of key search terms in context

• Wildcards are too wild

  "Tamecards" offer powerful generality without matching too much

• Retrieving and evaluating matching pages seems to take forever

  KWiCFinder works in the background, while you are doing something else.  It can save all downloaded documents matching your query in HTML (web) or text form for later reference or analysis

KWiCFinder Home
Screen Shots | Report Formats | Sample Search Reports
kfNgram Corpus Analysis Tool | Support