Literaturnachweis - Detailanzeige
Autor/inn/en | Crossley, Scott; Ocumpaugh, Jaclyn; Labrum, Matthew; Bradfield, Franklin; Dascalu, Mihai; Baker, Ryan S. |
---|---|
Titel | Modeling Math Identity and Math Success through Sentiment Analysis and Linguistic Features [Konferenzbericht] Paper presented at the International Conference on Educational Data Mining (EDM) (11th, Raleigh, NC, Jul 16-20, 2018). |
Quelle | (2018), (10 Seiten)
PDF als Volltext |
Sprache | englisch |
Dokumenttyp | gedruckt; online; Monographie |
Schlagwörter | Correlation; Speech Communication; Written Language; Mathematics Achievement; Self Concept; Natural Language Processing; Comparative Analysis; Mathematics Skills; Intelligent Tutoring Systems; Teaching Methods; Electronic Mail; Language Usage; Student Interests; Connected Discourse; Discourse Analysis; Mathematics Tests; Scores; Prediction; Blended Learning; Elementary School Students; Computational Linguistics; Texas Korrelation; Geschriebene Sprache; Mathmatics sikills; Mathmatics achievement; Mathematical ability; Mathematische Kompetenz; Selbstkonzept; Natürliche Sprache; Mathematics ability; Intelligentes Tutorsystem; Teaching method; Lehrmethode; Unterrichtsmethode; Elektronischer Briefkasten; Sprachgebrauch; Studieninteresse; Diskursanalyse; Vorhersage; Linguistics; Computerlinguistik |
Abstract | A number of studies have demonstrated strong links between students' language features (as found in spoken and written production) and their math performance. However, no studies have examined links between the students' language features and measures of their Math Identity. This project extends prior studies that use natural language processing (NLP) features to examine student language features and math performance, replicating their analyses. The study then uses NLP features to model students' Math Identity. Specifically, the study compares performance on basic math skills within an online math tutoring system to both student language (as captured in emails to a virtual pedagogical agent) and to survey measures of Math Identity (math self concept, interest, and value). Language features were analyzed by a number of NLP tools that extracted information related to text cohesion, lexical sophistication, and sentiment. The findings indicate weak to medium relationships between math scores and Math Identity and language features were able to predict a significant amount of the variance in each Math Identity variable and in math scores. The potential for these measures to inform interventions for students with lower Math Identity is discussed. [For the full proceedings, see ED593090.] (As Provided). |
Anmerkungen | International Educational Data Mining Society. e-mail: admin@educationaldatamining.org; Web site: http://www.educationaldatamining.org |
Erfasst von | ERIC (Education Resources Information Center), Washington, DC |
Update | 2020/1/01 |