Designing a Recommender System for Mobile Applications Focusing on Relative Importance Weights of Learner-related Variables

Abstract

To embrace why and how people learn and how to combine learner characteristics for recommending foreign language learning mobile applications (apps), this research presents a recommender system based on the relative importance weights of learner-related variables. In developing the system, 100 adult learners used 4-6 foreign language learning apps, resulting in 557 user-satisfaction data to calculate the relative importance of 14 learner-related variables in four categories: (a) demographic information, (b) motivational orientation for language learning (instrumental/integrative), (c) learning style, and (d) learning experience. The result showed that the model considering the relative importance weights of learner-related variables outperforms the dummy model in predicting users' satisfaction with the apps.