Abstract
The present study aims to investigate genre influence on the use and misuse of conjunctive adverbials (hereafter CAs) by compiling a learner corpus annotated with discoursal information on CAs. To do so, an online interface is constructed to collect and annotate data, and an annotating system for identifying the use and misuse of CAs is developed. The results show that genre difference has no impact on the use and misuse of CAs, but that there does exist a norm distribution of textual relations performed by CAs, indicating a preference preset in human cognition. Statistic analysis also shows that the proposed misuse patterns do significantly differ from one another in terms of appropriateness and necessity, ratifying the need to differentiate these misuse patterns. The results in the present study have three possible applications. First, the annotate data can serve as training data for developing technology that automatically diagnoses learner writing on the discoursal level. Second, the founding that textual relations performed by CAs form a distribution norm can be used as a principle to evaluate discoursal organization in learner writing. Lastly, the misuse framework not only identifies the location of misuse of CAs but also indicates direction for correction.
Translated title of the contribution | Diagnosing discoursal organization in learner writing via conjunctive adverbials |
---|---|
Original language | Chinese (Traditional) |
Pages | 310-322 |
Number of pages | 13 |
Publication status | Published - 2011 |
Event | 23rd Conference on Computational Linguistics and Speech Processing, ROCLING 2011 - Taipei, Taiwan Duration: 2011 Sep 8 → 2011 Sep 9 |
Conference
Conference | 23rd Conference on Computational Linguistics and Speech Processing, ROCLING 2011 |
---|---|
Country/Territory | Taiwan |
City | Taipei |
Period | 11-09-08 → 11-09-09 |
All Science Journal Classification (ASJC) codes
- Language and Linguistics
- Speech and Hearing