Keywords: Network Embedding, Heterogeneous Information Network
One of the challenges in mining information networks is the lack of intrinsic metric in representing nodes into a low dimensional space, which is essential in many mining tasks, such as anomaly detection, link prediction, and recommendation. Moreover, when coming to heterogeneous information networks, where nodes belong to different types and links represent different semantic meanings, it is even more challenging to represent nodes properly for a particular task. In this talk, I will introduce our recent progress of network embedding approaches that are designed for heterogeneous information networks, and discuss (1) how network embedding can be designed in unsupervised tasks, such as anomaly detection; (2) how to learn network embeddings when guidance is available, such as link prediction; and (3) how to design more complex embedding function when rich content information is available for nodes, such as recommendation. Our results on several application domains, including enterprise networking, social network, bibliographic data, and biomedical data, have demonstrated the superiority as well as the interpretability of these new methodologies.