一种采用偏最小二乘回归的情绪调节策略预测模型

摘要

A model of emotion regulation strategy using partial least squares regression is proposed based on personality traits to improve the situation that most researchers concern only learners' cognition, and construct a great amount of substantive digital learning resources, but neglect learners' affect in current personalized e-learning systems.The approach applies the principal component analysis and multiple regressions to extract some principle components in personality attributes, which are major influencing factors on emotion regulation strategy.Then the predictive model of emotion regulation strategy is demonstrated by an emotion chatting platform.Experimental results show that the explain ability of the emotion regulation strategy in personality characteristics reaches about 70％ and that the model can reveal the relation between personality characteristic and the method of emotion compensation.%针对目前个性化网络学习中重知识轻情感的现状,从学习者负面情绪调节入手,提出采用偏最小二乘回归方法,以采集的学习者的个性特征为自变量,负面情绪调节的策略特征作为应变量,建立学习者的情绪调节策略预测模型.通过实例验证,此模型具有较好的拟合预测能力.实验结果表明,个性对情绪调节策略的解释能力达70 %,揭示出学习者负面情绪补偿的方法与个性相关联.