The IEEE ICDM Workshop on Higher-Order Analytics of Large Graphs (HOALG) explores the challenges on high-order analytics of large graphs. Graphs are prominent in modeling the relationship of real-world entities, and its size is increased dramatically as the revolution of data collection endpoints. A motif, also known as a higher-order structure or graphlet, is considered as a fundamental "basic building block" of a large graph. Specifically, a motif is a small subgraph that appears frequently in a graph. A motif can help researchers to understand the significant, structural, or evolutionary design principles used to construct the large graphs. For example, a feed-forward loop motif is often used to study regulatory control mechanism in gene transcriptional networks. Hence, exact and approximate motif counting and discovery solutions have been developed intensively for different kinds of large graphs, e.g., heterogeneous information networks, typed graphs, uncertain graphs, dynamic and temporal graphs. New computing architectures and hardwares have been applied to make it more efficient and scalable while maintaining the mining effectiveness. Also, researchers used motifs extensively for higher-order graph analytics solutions, such as graph clustering, ranking, embedding, visualization, link prediction, recommendation, deep learning model design, fraud detection, community search, cliques and densest subgraph discovery. New insights or better analytic effectiveness are often obtained when enabling the higher-order semantics with the help of motifs. 

This workshop has a number of objectives:

• Avenue for Presenting Research: Provide a forum for presenting research in this emerging area of motif discoveries and higher-order graph analytics.

• Platform for Discussions: Provide a platform for researchers interested in this area to engage in discussions on how this emerging area could shape up in the future.

• Cross-pollination: Encourage computer science researchers from different domains (e.g., data mining and management, computer systems, and artificial intelligence), to share their perspectives and visions on this area, and help computing researchers to realize potential for cross-disciplinary approaches in this area to eliminate any systemic blind spots.