Coevolution and Many-Objective Search Optimization in Multilayer Social Networks

多层社交网络中的协同进化和多目标搜索优化

• 批准号: RGPIN-2022-04017
• 负责人: MoradianZadeh, Pooya
• 金额: $ 1.82万
• 依托单位: University of Windsor
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2022
• 资助国家: 加拿大
• 起止时间: 2022-01-01 至 2023-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=750372
• 关键词: Coevolution; Many; Objective; Search; Optimization

Nowadays, a vast amount of social data is available, which can be utilized by social network analysis to model and study the structure and dynamics of real-life complex systems such as social and economic systems. It provides a wide range of opportunities for industries and decision-makers to optimize their procedures and increase efficiency across a broad area of applications, from marketing and campaign analysis to user-centric recommendation systems and from dynamic resource allocation to energy management and urban planning. Social networks are mostly modeled as graph structures where the social actors are represented by the nodes and the edges, representing interactions or relationships between them. However, social actors in the most complex and social systems usually are members of multiple networks simultaneously, and more than one kind of relationship exists between them. For example, a person may be linked to someone in a friendship network while they are also connected in a professional network. Having a membership in multiple social networks simultaneously has an enormous impact on their actions and decision-making process, highlighting the need for multi-layer social network analysis, where each layer represents the interactions and behavior of the actors in different social contexts and environments (e.g.,, friendship network). In recent years, much attention has been paid to the structure of these social networks. However, there are still many open problems in the field due to their complex and dynamic nature. This proposal focuses on developing effective and efficient solutions to explore the co-evolution of dynamic multi-layer social networks and study their behaviors over time using computational intelligence and social network analysis techniques. We build a multi-layer dual-inheritance evolutionary framework that utilizes the extracted knowledge from both the networks' structures and individual behavior to track the co-evolution. We also study social influence in these networks and define metrics and methods to measure it.   Additionally, we will use the proposed framework to study the behavior of real-life social networks and investigate their underlying patterns. We will further study the problem of many-objective search optimization in these co-evolving networks and address critical issues, challenges, and opportunities associated with their heterogeneous, large-scale, and complex structures.

如今，大量的社会数据可以被社会网络分析用于建模和研究现实生活中的复杂系统，如社会和经济系统的结构和动力学。从市场营销和活动分析到以用户为中心的推荐系统，从动态资源分配到能源管理和城市规划，它为行业和决策者提供了广泛的机会，以优化其程序和提高应用程序的效率。社会网络大多被建模为图结构，其中社会参与者由节点和边来表示，表示它们之间的交互或关系。然而，在最复杂的社会系统中，社会行为者通常同时是多个网络的成员，并且他们之间存在着多种关系。例如，一个人可能被链接到友谊网络中的某人，而他们也在职业网络中连接。同时拥有多个社交网络的成员资格对他们的行动和决策过程具有巨大的影响，突出了多层社交网络分析的必要性，其中每一层代表参与者在不同社交背景和环境(例如，友谊网络)中的交互和行为。近年来，这些社交网络的结构受到了人们的广泛关注。然而，由于其复杂性和动态性，该领域仍然存在许多悬而未决的问题。这项建议的重点是开发有效和高效的解决方案，以探索动态多层社会网络的协同进化，并使用计算智能和社会网络分析技术研究它们随时间的行为。我们构建了一个多层双继承进化框架，该框架利用从网络结构和个体行为中提取的知识来跟踪共同进化。我们还研究了这些网络中的社会影响力，并定义了衡量它的指标和方法。此外，我们还将使用提出的框架来研究现实生活中的社交网络的行为，并调查其潜在模式。我们将进一步研究这些协同进化网络中的多目标搜索优化问题，并解决与其异质、大规模和复杂结构相关的关键问题、挑战和机会。