Polish search engine uses word inflections
January 31st, 2008
Morfeo is the name of the search engine technology used by one of the largest Polish search engines www.szukaj.onet.pl, and also by search engines in the Czech Republic and Slovakia.
It differs from the English-based technologies in the way it treats word morphology. Polish is highly inflected and retains the Old Slavic case system with seven cases for nouns, pronouns, and adjectives, which means that depending on the case of the word, the word ending changes.
Morfeo allows someone searching the word “zdrowie” (health) to find pages which include inflected versions of the search word, such as “zdrowia”, “zdrowiu” etc., which may be equally relevant for the search, but disregarded by English-based search engines. However, an exact match of the word case is ranked higher in the search results than the inflected versions. Naturally, this is not the only criterion for search relevance.
The technology is being developed by an international team concentrated around the Charles University in Prague and Jagiellonian University in Cracow.
