|
EPSA [10]-10This is a Sub-Experiment for EPSA [1]-n and EPSA [33]-n. The specification of the intended dependency interpretation is prose-based. The Examples in EPSA [10]-5 and EPSA [10]-10 are identical. The two LGs in EPSA [10]-5 and [10]-10are BVA(mittu-no kyuudan, soko/asoko) 'BVA(three ball clubs, it)' and BVA(subete-no kyuudan, soko/asoko) 'BVA(every ball club, it)'. DesignExamplesResultsEffects of informant classification on a *Schema-based predictionInformant list for [10]-10: The informant classification is based on the result of [3]-7, considering Lexical group 3 (soitu-tati vs. soitu) and Lexical group 4 (sono 2-sya vs. soko). We consider only those informants whose %(Y) on B is 0% and whose %(Y) on A is 50% or higher in [3]-7, both with Lexical group 3 and Lexical group 4. [10]-10 Summary, based on the above informant list [10]-10 Lexical-group-based, based on the above informant list [10]-10 Summary as of Oct. 7, 2015 Informant list for [10]-10: The informant classification is based on the result of [10]-5, considering both of the LGs. (There is only one SG in [10]-5.) We consider only those informants whose %(Y) on B is 0% and whose %(Y) on A is 25% or higher in [10]-5, as of Oct. 7, 2015 [10]-10 Summary, based on the above informant list, as of Oct. 7, 2015 The one-sentence-at-a-time test type vs. the three-sentences-at-a-time test type[10]-10 One sentence at a time [10]-10 Three sentences at a time Informant list for [10]-10: The informant classification is based on the result of [3]-7, considering Lexical group 3 (soitu-tati vs. soitu) and Lexical group 4 (sono 2-sya vs. soko). We consider only those informants whose %(Y) on B is 0% and whose %(Y) on A is 50% or higher in [3]-7, both with Lexical group 3 and Lexical group 4. [10]-10 One sentence at a time, based on the above informant list [10]-10 Three sentences at a time, based on the above informant list Raw data |