Abstract:

Two major problems of the Digital Age are on the one hand decreasing half-life of knowledge caused by increasing progress and penetration of technology, and on the other hand multifaceted demands of Industry 4.0 regarding time and location-independent, extra-occupational education just in time and on demand. More and more, non-formal and informal learning within distance e-learning scenarios become part of lifelong learning processes. To overcome the Web 2.0 caused information overload that hampers finding, filtering, evaluating and processing appropriate learning material, social learning strategies like Collaborative and Cooperative Learning, Social Constructivism, Social Learning, and Connectivism offer a solution for this problem.

In order to connect people within social learning environments on the basis of individual context information, this paper presents the “Dharma” (Dynamic, heuristic & adaptive peer recommendation in social learning environments) algorithm. Initiated by a person, who for example wants to start collaborative working, cooperative authoring, or just needs help in case of a problem of comprehension, Dharma recommends suitable peers based on thematic (experiences, learning successes, authoring activities) and social competence (online activity, community ratings), motivation (same interests and learning targets), and compatibility between the users (same languages, social connection, preferences, and situation). Learning groups can be generated automatically by inviting a default number of top ranked users. In manual mode, results are visualized as individual radar charts of each user, in which each analysed attribute is reflected by an axis. In this way, the person who initiated the group forming process can easily compare potential peers. By manipulating weight parameters for each attribute, the rankings can also be rearranged based on individual preferences. In contrast to the state of the art, Dharma includes a monitoring mechanism that dynamically invites new users if there’s low interaction activity in the learning group, or conversation topic has been changed.

To motivate cooperativeness and to avoid misuse of this helping functionality, karma points are used as a virtual currency. By raising the social competences, for instance by being an active member who is regularly online and supports other user’s learning processes, and the thematic competences, for instance by independently acquiring new knowledge that may useful for others, users can earn karma points that in turn can be payed to initiate a group formation by Dharma. Selected peers who discussed actively and achieved positive ratings by the other participants get more karma points than someone who is not available and not participating in the group.

A fully functional prototype of the Dharma algorithm is embedded into the Web 3.0 learning and teaching platform Wiki-Learnia. It is shown how a social learning system should be designed in order to meet the requirements of innovative social distance e-learning. To evaluate the overall approach including fitting peer recommendation and the virtual currency principle, Wiki-Learnia will be prepared for application in the Junior Study program of the University of Rostock. More than 250 people that are distributed throughout Germany participate each semester in the video-based online study course for pupils. Hence, it forms an ideal testbed for the described solution.

Keywords:

CSCL, peer finding, group forming, distance learning, learning network.