Focusing on applications for analyzing learner language which evaluate semantic appropri- ateness and accuracy, we collect data from a task which models some aspects of interac- tion, namely a picture description task (PDT). We parse responses to the PDT into depen- dency graphs with an an off-the-shelf parser, then use a decision tree to classify sentences into syntactic types and extract the logical sub- ject, verb, and object, finding 92% accuracy in such extraction. The specific goal in this paper is to examine the challenges involved in ex- tracting these simple semantic representations from interactive learner sentences. 1 Motivation While there is much current work on analyzing learner language, it usually focuses on grammati- cal error detection and correction (e.g., Dale et al., 2...

Click Here to View the Paper In XML Format