Skip to main content.

Big Data Analysis using Graphs and MLNs

Graphs and MLNs provide a powerful foundation for analyzing large, complex datasets where entities and their relationships span multiple types, contexts, or data sources. By modeling data as interconnected layers, MLNs capture richer structural and semantic information than single-layer graphs, enabling deeper insights across diverse domains such as social networks, healthcare, cybersecurity, and scientific data.

  1. Centrality

    2. Centrality measures identify the most important or influential nodes within a graph or MLN. Our research focuses on novel approaches to compute various types of centrality measures (e.g. Closeness Centrality, Betweenness Centrality) directly on MLNs using a decoupling-based approach.

  2. Community Detection

    3. Community detection groups nodes into clusters with densely connected internal relationships. MLNs enable communities to be discovered not only within individual layers but also across them.

  3. Substructure Discovery

    4. Substructure discovery focuses on discovering frequent, interesting patterns in graph data. Our work is focused on developing scalable approaches to perform substructure discovery on partitioned single graphs and MLNs.