Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Combinational Nonuniform Timeslicing of Dynamic Networks

Seokweon Jung, DongHwa Shin, Hyeon Jeon, Jinwook Seo

TL;DR

The paper addresses the challenge of choosing optimal nonuniform time slices for dynamic networks, where uniform intervals fail to capture density and structural changes. It introduces a combinational approach that first applies Visual Complexity–based slicing (VS) to create fine-grained segments and then aggregates them with Jaccard similarity–based slicing (JS) to form nonuniform snapshots, calibrating to a weekly interval to produce 17 snapshots. On a Rugby Tweets dataset, the hybrid method yields intermediate snapshot density that preserves evolving patterns and offers more detailed temporal structure than either baseline method alone. This fusion of data mining and visualization techniques enhances interpretability and pattern discovery in dynamic networks, suggesting practical improvements for temporal network analysis.

Abstract

Dynamic networks represent the complex and evolving interrelationships between real-world entities. Given the scale and variability of these networks, finding an optimal slicing interval is essential for meaningful analysis. Nonuniform timeslicing, which adapts to density changes within the network, is drawing attention as a solution to this problem. In this research, we categorized existing algorithms into two domains -- data mining and visualization -- according to their approach to the problem. Data mining approach focuses on capturing temporal patterns of dynamic networks, while visualization approach emphasizes lessening the burden of analysis. We then introduce a novel nonuniform timeslicing method that synthesizes the strengths of both approaches, demonstrating its efficacy with a real-world data. The findings suggest that combining the two approaches offers the potential for more effective network analysis.

Combinational Nonuniform Timeslicing of Dynamic Networks

TL;DR

The paper addresses the challenge of choosing optimal nonuniform time slices for dynamic networks, where uniform intervals fail to capture density and structural changes. It introduces a combinational approach that first applies Visual Complexity–based slicing (VS) to create fine-grained segments and then aggregates them with Jaccard similarity–based slicing (JS) to form nonuniform snapshots, calibrating to a weekly interval to produce 17 snapshots. On a Rugby Tweets dataset, the hybrid method yields intermediate snapshot density that preserves evolving patterns and offers more detailed temporal structure than either baseline method alone. This fusion of data mining and visualization techniques enhances interpretability and pattern discovery in dynamic networks, suggesting practical improvements for temporal network analysis.

Abstract

Dynamic networks represent the complex and evolving interrelationships between real-world entities. Given the scale and variability of these networks, finding an optimal slicing interval is essential for meaningful analysis. Nonuniform timeslicing, which adapts to density changes within the network, is drawing attention as a solution to this problem. In this research, we categorized existing algorithms into two domains -- data mining and visualization -- according to their approach to the problem. Data mining approach focuses on capturing temporal patterns of dynamic networks, while visualization approach emphasizes lessening the burden of analysis. We then introduce a novel nonuniform timeslicing method that synthesizes the strengths of both approaches, demonstrating its efficacy with a real-world data. The findings suggest that combining the two approaches offers the potential for more effective network analysis.
Paper Structure (4 sections, 1 figure)

This paper contains 4 sections, 1 figure.

Table of Contents

  1. Introduction
  2. Combinational Nonuniform Timeslicing
  3. Evaluation
  4. Conclusion and Future Work

Figures (1)

  • Figure 1: Result of the combinational nonuniform timeslicing utilizing both Jaccard similarity-based and visual complexity-based slicing method. The noise at the beginning reduced as the number of events in each snapshot increased while showing significant changes over time (green box). Snapshots at the end now clearly visualize the final evolution of the network, revealing a hidden pattern (red box).