Descriptive vs. Inferential Community Detection in Networks: Pitfalls, Myths and Half-Truths

Community detection is an important field of network science, but there is a gap between state-of-the-art and practical methods. The Elements attempts to address this by dividing methods into descriptive and inferential goals, providing insights into formation mechanisms and separating structure from noise.

Format: Paperback / softback
Length: 75 pages
Publication date: 31 August 2023
Publisher: Cambridge University Press

Community detection is a crucial and highly sought-after field within the realm of network science, garnering substantial attention over the past several decades. Its primary objective is to automate the division of a network into its fundamental building blocks, aiming to provide a concise summary of its overarching large-scale structure. Despite its immense significance and widespread adoption across various domains, there exists a notable discrepancy between the state-of-the-art methods and the techniques that are actually employed in practice. This gap is particularly evident in diverse fields such as social networks, biological systems, and complex networks.

To address this discrepancy, The Elements takes a novel approach by categorizing existing methods based on whether they have a descriptive or inferential goal. Descriptive methods focus on identifying patterns in networks based on context-dependent notions of community structure. On the other hand, inferential methods articulate a precise generative model and strive to fit it to data, offering insights into formation mechanisms and distinguishing between structure and noise.

By adopting this classification system, The Elements aims to bridge the gap between theory and practice, providing a comprehensive resource for researchers and practitioners alike. The book offers a detailed exploration of various community detection algorithms, their strengths, limitations, and applications, making it an invaluable tool for those seeking to understand and analyze complex networks. Moreover, the title is available as open access on Cambridge Core, ensuring widespread accessibility and facilitating interdisciplinary research.

In conclusion, community detection plays a pivotal role in advancing our understanding of complex networks and their dynamics. The Elements provides a valuable contribution to this field by categorizing existing methods, offering insights into formation mechanisms, and facilitating interdisciplinary research. Its open access availability on Cambridge Core further enhances its accessibility and promotes the exchange of knowledge and ideas among researchers and practitioners.

Weight: 135g
ISBN-13: 9781009113007