Comparing Different Approaches to Treat Translation Ambiguity in CLIR: Structured Queries vs. Target Co-occurrence Based Selecti

—Two main problems in Cross­language Information  Retrieval are translation selection and the treatment of out­of­ vocabulary  terms.  In this  paper,  we  will  be  focusing  on  the  problem concerning the translation selection. Structured queries  and target co­occurrence­based methods seem to be the most  appropriate approaches when parallel corpora are not available.  However,   there   is   no   comparative   study.   In   this   paper   we  compare the results obtained using each of the aforementioned  methods, we specify the weaknesses of each method, and finally  we propose a hybrid method to combine both. In terms of mean  average   precision,   results   for   Basque­English   cross­lingual  retrieval show that structured queries are the best approach  both with long queries and short queries. 

• Type: Conference
• Year: 2009
• Where: DEXAW
• Authors: Xabier Saralegi, Maddalen Lopez de Lacalle
• Keywords: Database, DEXAW 2009, Problem concerning the translation selection. Structured queries And target co­occurrence­based methods seem to be the most Appropriate approaches when parallel corpora are not available. , Retrieval are translation selection and the treatment of out­of­ Vocabulary Terms, We propose a hybrid method to combine both. In terms of mean Average Precision