This work focuses on estimating the impact of produced arguments on the discussion's upcoming development. We sought to define a feature upon which the productivity of non-convergent discussions can be predicted. We rely on the Bakhtinian notion of responsiveness, the degree to which a speaker embeds or builds on the arguments of an interlocutor. We demonstrate the potential of this feature, using a corpus of 10,000 threads extracted from Reddit's 'Change My View' forum. Utilizing Synthetic Minority Oversampling Technique, we experimented with several supervised machine learning algorithms, each of which drew on a different set of selected features. The performance of the obtained models was evaluated through repeated stratified 10-fold cross-validation. Our preliminary results are encouraging: responsiveness contributes to accurate prediction of discussion productivity.
|Title of host publication
|ISLS Annual Meeting 2023
|Subtitle of host publication
|Building Knowledge and Sustaining our Community - 16th International Conference on Computer-Supported Collaborative Learning, CSCL 2023 - Proceedings
|Crina Damsa, Marcela Borge, Elizabeth Koh, Marcelo Worsley
|International Society of the Learning Sciences (ISLS)
|Number of pages
|Published - 2023
|16th International Conference on Computer-Supported Collaborative Learning, CSCL 2023 - Montreal, Canada
Duration: 10 Jun 2023 → 15 Jun 2023
|Computer-Supported Collaborative Learning Conference, CSCL
|16th International Conference on Computer-Supported Collaborative Learning, CSCL 2023
|10/06/23 → 15/06/23
Bibliographical notePublisher Copyright:
© 2023 International Society of the Learning Sciences (ISLS). All rights reserved.