MiLQ is a benchmark of mixed-language (code-switched) search queries created by bilingual speakers for evaluating Information Retrieval with mixed-language queries. It provides query versions where English expressions are embedded within native-language structures. This work is derived from The CLEF Test Suite for the CLEF 2000-2003 Campaigns available in the ELRA Catalogue (ELRA-E0008). Languages covered: Swahili (SW), Somali (SO), Finnish (FI), German (DE), French (FR), Chinese (ZH), Persian (FA), Russian (RU) Format: • Title: short keyword-style query per topic • Description: single-sentence query per topic. Statistics: • Swahili (SW): 151 Title + 151 Description (302 queries). • Somali (SO): 151 Title + 151 Description (302 queries). • Finnish (FI): 151 Title + 151 Description (302 queries). • German (DE): 151 Title + 151 Description (302 queries). • French (FR): 151 Title + 151 Description (302 queries). • Chinese (ZH): 47 Title + 47 Description (94 queries). • Persian (FA): 45 Title + 45 Description (90 queries). • Russian (RU): 44 Title + 44 Description (88 queries).

Creator(s)

-

Distributor(s)

ELRA

Right Holder(s)

-