Link Prediction and Analysis on Knowledge Graphs - RMGS
DescriptionBy owning the great power in modeling the relationships and interactions between different units, network, or graph, analysis has a wide range of applications both in researchers and industry, e.g. flow control, computer network, database management, and bioinformatics. There are various knowledge networks in real world. For example, biological networks, social networks, traffic networks, citation networks and so on. In a biological network, the node can refer to a gene, a protein, or a species in the microbial community, and edge captures the relationship between the nodes, like how similar is the expression profile of two genes, cooperation relationship, or regulatory relationship.Properly representation and analysis can provide insights into the structure of the network, therefore, give useful knowledge. Great efforts have been made to develop methods and tools for network analyzation. Researcher finds that many real-world networks can be regarded as complex networks. Such networks preserve property, like small-world and scale-free and usually in large scale. However, the boosted size of real-world networks makes the high demand for memory and time for network storage and network analysis, which limits the usage.In this project we aim to survey the existing methods and models for network embedding and network analysis, including linkage prediction, nodes classification, visualization to recover essential information in the network. We will also propose multiple models for network embedding and multiple methods for link prediction and nodes classification.
|Effective start/end date||1/03/20 → …|