Recently Facebook supports the social feature of communication, and further offering individual profiles to share with friends, and applying interactive platform to discuss the specific issues with each other. In a collaborative learning environment, a learning companion could assist learners to understand the course content, share knowledge, arouse creativity and facilitate the effects of collaborative learning. Meanwhile, the learning companions of different learning styles may affect the learning effect for the leaner. For this reason, this study develops an individual learning companion recommendation system on Facebook, and supports mobile collaborative learning. The system automatically collects information from friends' profiles that can provide data about their learning needs, such as interests, location, learning styles and professional abilities. In addition, this system utilizes an Artificial Bee Colony algorithm to search for the optimal learning companion. The results indicate that the proposed method outperforms other approaches and improves the accuracy of the search process.
