| 研究課題/領域番号 |
20H04288
|
|---|
| 研究機関 | 東京工業大学 |
|---|
研究代表者 |
Barrat Alain 東京工業大学, 科学技術創成研究院, 特任教授 (10867287)
|
|---|
| 研究分担者 |
Holme Petter 東京工業大学, 科学技術創成研究院, 特任教授 (50802352)
Jusup Marko 東京工業大学, 科学技術創成研究院, 特任助教 (60762713)
村田 剛志 東京工業大学, 情報理工学院, 教授 (90242289)
|
|---|
| 研究期間 (年度) |
2020-04-01 – 2024-03-31
|
| キーワード | Complex networks / temporal networks / epidemic processes / social contagion |
|---|
| 研究実績の概要 |
For data represented as networks in particular, several methods have been proposed to extract a ”network backbone”, i.e., the set of most important links. However, the question of how the resulting compressed views of the data can effectively be used has not been tackled. We have addressed this issue by putting forward and exploring several systematic procedures to build surrogate data from various kinds of temporal network backbones. In particular, we have explored how much information about the original data need to be retained alongside the backbone so that the surrogate data can be used in data-driven numerical simulations of spreading processes. In another project related to the issue of the representation of network data, we have proposed a node embedding technique aimed at providing low-dimensional feature vectors that are informative of dynamical processes occurring over temporal networks - rather than of the network structure itself. we have moreover considered the potential impact of digital contact tracing within the containment measures of the Covid-19 pandemic. Several countries have thus deployed apps that can automatically detect contacts, but the efficiency of such measures is unknown. We have thus investigated this point by several modelling approaches, showing that the efficiency grows only quadratically with the app adoption, but that no critical app adoption threshold exists: any increase in adoption increases the efficiency of the measure.
|
| 現在までの達成度 (区分) |
現在までの達成度 (区分)
2: おおむね順調に進展している
理由
We have started to investigate a possible definition of “temporal rich club” on temporally resolved datasets of various origins and describing different systems: from face-to-face interactions to infrastructure networks to brain networks. A first characterization seems to reveal interesting patterns. We need to confirm the relevance of these patterns by considering randomized versions of the same datasets. We also need to explore the emerging temporal rich clubs to relate them to the existing metadata and characterize them.
We have implemented the new representation of social ties and now need to explore systematically various types of perturbations in both data and tunable models, and the ability to detect them.
Finally, we have considered the case of a single committed minority in a well-known model for the emergence of social norms. We plan to now consider the case of competing minorities.
|
| 今後の研究の推進方策 |
We will develop new methods and tools to deal with network data, in particular concerning temporal networks, in two main research directions.
On the one hand, we will define and investigate new types of relevant structures in temporal network data, such as the 'temporal rich club': we will define a quantity that describes the tendency of well connected nodes of the network to be connected together in a simultaneous fashion (generalizing hence the usual rich club coefficient for static networks, which does not take into account temporality). We will study data sets of different types, as well as models of temporal networks, and check whether they exhibit such temporal rich clubs.
On the other hand, we will consider a new representation of social ties built from temporal network data that takes into account the interdependency of social relationships. Using a series of temporal network models with tunable properties, and tailored perturbations of these networks, we will investigate the ability of this representation to detect perturbations in a social system. We will moreover use this representation to propose new ways of modeling social contagion processes in a network.
|
