Corpus description. The IFCASL corpus is a bilingual speech corpus for French and German language learners. It was designed in order to allow an in-depth analysis of both segmental and prosodic aspects of the non-native production of these languages. The corpus was recorded by fifty French learners of German and forty German learners of French in their native and second languages. The non-native speakers were classified by L1 teachers in three categories: beginners, intermediate and advanced. The corpus consists of four sets of sentences, corresponding to different speaking conditions: (1) reading sentences (about 30 sentences, referred to as SR sentences);
Appears in 6 contracts
Samples: www.isca-archive.org, www.ifcasl.org, citeseerx.ist.psu.edu