Ambiguity of Queries and the Challenges for Query Language Detection

In this paper, a sample set of 510 simple searches from the TEL action log 2009 is analyzed for query content and query language. More than half of the queries are for named entities, which has consequences for query language disambiguation. A manual identification of query language finds that often a definite language cannot be determined, because many named entities are not translated. Problems and challenges for query category and language identification are discussed. Further analysis shows that IP address and interface language are not very strong indicators for determining the query language.

• Type: Conference
• Year: 2010
• Where: CLEF
• Authors: Juliane Stiller, Maria Gäde, Vivien Petras
• Keywords: CLEF 2010, Information Technology, Query Content, Query Language, Query Language Disambiguation