Multiple Admissibility in Language Learning: Judging Grammatically Using Unlabeled Data

摘要

We present our work on the problem of detection Multiple Admissibility (MA) in language learning. Multiple Admissibility occurs when more than one grammatical form of a word fits syntactically and semantically in a given context. In second-language education—in particular, in intelligent tutoring systems/computer-aided language learning (ITS/CALL), systems generate exercises automatically. MA implies that multiple alternative answers are possible. We treat the problem as a grammatical-ity judgement task. We train a neural network with an objective to label sentences as grammatical or ungrammatical, using a 'simulated learner corpus': a dataset with correct text and with artificial errors, generated automatically. While MA occurs commonly in many languages, this paper focuses on learning Russian. We present a detailed classification of the types of constructions in Russian, in which MA is possible, and evaluate the model using a test set built from answers provided by users of the Revita language learning system.