Graphlet types and application domains: a review

Abstract

Graphlets-small, connected, non-isomorphic induced subgraphs-are powerful tools for capturing local structural patterns in complex networks. Graphlets help to detect anomalies or communities, compare complex networks, and make better recommendations. While much of the literature has focused on improving the efficiency of graphlet enumeration, limited attention has been paid to understanding how different types of graphlets (e.g., triangles, 4-cycles, cliques, stars) are applied across diverse domains. This paper introduces a conceptual framework that categorizes graphlet utility according to structural, functional, and frequency-based dimensions and also summarizes graphlet measures such as Graphlet Degree Vector, Graphlet Degree Distribution, and Graphlet Frequency Distribution, and analyzes how these tools support domain-specific tasks. The study offers a structured overview of graphlet usage across fields, including bioinformatics, social network analysis, and anomaly detection, highlighting their practical relevance and encouraging further context-aware research.