Error description and standardization of Modern Greek based on a context-sensitive grammar: Implications for the teaching of mother tongue
Περίληψη
The current study presents an analysis regarding the standardization of grammatical phenomena as well as corresponding templates for the description of grammatical phenomena / errors in order to develop an advanced, innovative and state-of-the-art computing environment for the creation of digital educational games for students. These templates were based on studies on the definition of linguistic error and focus on cases of common errors and deviations in the field of morphology / inflection, spelling and syntax. The implementation formalism used is Mnemosyne’s Kanon, a complete complex natural language processing system that is applied both for information retrieval and extraction in free texts. At the same time, the grammar checker that has been developed based on the aforementioned process, was put to the test during the correction of authentic texts of primary school students (N = 100), which were produced in the context of teaching scenarios in order to investigate its usability and contribution to their linguistic competence. Main findings are that a context sensitive grammar can address the problem of correcting spelling mistakes that result in legitimate words. Moreover, it was found that the proposed formalism, the templates and, consequently, the grammar checker were overall functional in detecting most of the participants’ usual errors.
Λεπτομέρειες άρθρου
- Ενότητα
- Articles