Effectiveness of Linguistic and Learner Features for Listenability Measurement Using a Decision Tree Classifier
Year: 2017
Type: article
Abstract: As the ease of grasping the contents of listening material influences learners' motivation and learning outcome, language teachers need to choose materials appropriate for the proficiency of their learners. This heavy task has been addressed by using a traditional readability measurement method to develop an automatic measurement method of the ease of listening comprehension using linear regression analysis for listening materials. Because machine learning such as decision tree classification can properly handle different types of features, recent readability measurement methods use classification approaches such as a decision tree. Then, we proposed a measurement method using decision tree classification for linguistic features of listening materials as well as learner features of listening proficiency. The experimental results showed that the accuracy of our method (47.0%) was better than the baseline accuracy (25.2%), and that the listening test score and visiting experience in English speaking areas among the learner features were discriminative for the measurement accuracy.
DOI: 10.12937/ejsise.16.7
Authors: Katsunori Kotani, Takehiko Yoshimi
Institutions: Kansai Gaidai University, Ryukoku University
Language: English
Subfield: Linguistics and Language
Field: Social Sciences
Domain: Social Sciences
SDG: Quality education
Open Access status: diamond
OpenAlex ID: W2753815938
