Dictionaries and Tolerant Retrieval
Dictionaries and Tolerant Retrieval
Dictionaries and Tolerant Retrieval
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
Introduction to Information <strong>Retrieval</strong>Sec. 3.4Soundex• Soundex is the classic algorithm, provided by mostdatabases (Oracle, Microsoft, …)• How useful is soundex?• Not very –for information retrieval• Okay for “high recall” tasks (e.g., Interpol), thoughbiased to names of certain nationalities• Zobel <strong>and</strong> Dart (1996) show that other algorithms forphonetic matching perform much better in thecontext of IR45