Online learner emotional analysis based on big dataset of online learning forum

Autor: Xiaoli Yang, Changri Luo, Baohua Huang, Xinhua Zhang, Tingting He, Zizhou Lu
Rok vydání: 2017
Předmět:
Zdroj: CISP-BMEI
DOI: 10.1109/cisp-bmei.2017.8302316
Popis: The spatial and temporal separation characteristics of online learning not only make learners lack the presence of classroom teaching, but also make learners feel lonely during learning. The emotional analysis of learners is helpful for the Learning support services provider to online learning support services. In the online learning forum, the Learner's post information is the direct carrier of learner's emotional sustenance. Therefore, the analysis of learners' emotion is the analysis of the learners' posts in the online learning forum. This paper collects 1.1 million posts of online corpus (Dataset DI and DII). Based on the support of emotional dictionary, this paper uses the data mining method to analyze the learners' emotion. It is found that: (a) the proportion of negative emotions of online academic education and online training learners is 6.35% and 5.12% respectively, but the positive emotion is 27.02% and 90.91% respectively; (b) The proportion of negative emotions in each test course of DI is higher, which reflects the greater learning pressure; (c) The learning support service forum is the most number of concerned on and negative emotion, indicating that learning support services to be further strengthened.
Databáze: OpenAIRE