الفهرس | Only 14 pages are availabe for public view |
Abstract The growth of social networks has lately attracted both academic and industrial researchers to study the ties between people, and how the social networks evolve with time. Social networks like Facebook, Twitter and Flickr require efficient and accurate methods to recommend friends to their users in the network. Link prediction approaches have been devised to recommend friends, or collaborators in academic networks and predict likelihood of future links in general. Edges and published content of the network are the features used for prediction. Two main link prediction approaches are used to utilize these features: Score-based approaches and machine learning approaches. Score-based approaches calculate similarity between users and determine the likelihood of forming future links, based on a formula using the network features, while machine learning approaches treat the prediction process as a classification problem. The classifier predicts the class of each edge whether it exists or does not exist. Machine learning approaches have the benefit of adding all similarity indices needed as the feature set fed to the classifier. While in score-based approach when we used multiple features with associated weights, the performance was sensitive to the values of such weights. Even though link prediction is a popular research area, but most of the work done in this area misses many of the features that play an important role in link formation |