This research provides a new model for team formation problems for Impromptu task activities in geo-social networks, called Geo-team formation. The team formation problem (TFP) is the process of dedicating the users from Social Networks to activities as teams in a collaborative functioning environment for an effective outcome based on their skills. It was proven to be an NP-hard problem. Impromptu activities deployment requires users with required skills who are socially close to each other and spatially close to the activity location. Most existing researches tackle the geo-team formation problem as a single social constraint or skills constraints query while optimizing the spatial closeness. To cover this gap, we present a model that efficiently narrows down the search space and then applies the required constraint. Efficient processing of the geo-team formation model is challenging considering (1) required skills of users, (2) weight of user skills, (3) the maximum contribution of users skills, (4) social cohesiveness between users, and (5) spatial closeness, which needs to be carefully examined and made timely invitations. The weight of user's skills helps the model dedicate users with high expertise for each team's required activities. In this approach, the time and the search cost of the team formation process are efficiently restricted. Our preliminary results on real-world datasets determine that the proposed algorithms can efficiently apply the geo-team formation model under various parameter settings. the results confirm the effectiveness of the proposed method.
Article ID: 2021G11
Publisher: Canadian Artificial Intelligence Association