Thematic Network Modelling for Digital Ecosystems

研究内容从传统传播方式向数字传播方式的转变带来了海量的信息传播，几乎没有监管，造成了数字生态系统和近年来日益及时的跨学科话题。例如，数字生态系统可以包括社交媒体平台和数字市场。因此，理解这些制度，并提出适当的政策，最近变得越来越突出。目前的计算技术很少从系统层面的角度关注每个个人/组织/公司的行为在生态中所扮演的内在角色，例如在线社交媒体中的信息创造者、宣传者和倾听者，或者数字市场中每个利益相关者所拥有的关键/外围排名。研究的目的和目标本项目旨在发展数字生态系统中相互作用动力学的计算分析、模拟和干预，寻求回答以下关键问题：1.如何评估每个个人/组织的固有角色。以一种系统可解释的方式发挥作用。我们能否通过角色结构和行为模拟对多面性群体进行有意义的比较。为了改善数字生态系统，基于以上分析，可以提出哪些干预途径？这可能会对推荐算法有所帮助。上述方法的可扩展性限制是什么，我们能否缓解这些限制，以便能够分析不断增加的数据大小(例如，跨大人口样本或不同地理位置)？研究方法的新颖性本研究的主要新颖性是应用网络结构技术来评估数字生态系统中系统级的集体行为特征，并相应地提出相关的干预方法，特别是与推荐算法相关。潜在的影响、应用和好处将复杂的网络结构引入数字生态系统，可以为广泛的用户交互动态提供直观的系统级评估。可能会探索适当的干预方法，以潜在地改善数字注册和政策。研究与汇款的关系本研究与EPSRC的任务在方法论和目标上都密切相关。复杂网络建模作为主要方法论，在很大程度上依赖于数学科学。目标是改善数字生态系统，包括数字经济和信息通信技术。外部合作伙伴-StateUP(https://stateup.co/nebula/)-StateUp是一家初创公司，站在工业和学术界的十字路口，致力于公共部门的采购技术。它将为我提供实习机会和数据集。具体地说，我可以帮助建立星云，这是一个由State Up管理的储存库，可以用来向寻求解决各种规模的社会问题的公共机构提供建议。在此过程中，我还将独家访问认可的公司简介数据集。

The contect of the researchThe shift from traditional communication methods towards digital ones brings about massive information dissemination with little to no regulation, making difgital ecosystems and increasingly timely interdisciplinary topic in recent years. Digital ecosystems can include, for example social media platforms and digital markets. Understanding such systems and proposign appropriate policies have therefore become more and more prominent recently. Current computational techniques rarely pay attention to the inherent role each individual/organisation/company's behaviour plays in the ecology from a system-level perspective, such as information creator, propagandist and listener in the online social media, or crucial/peripheral rank each stakeholder posesesses in a digital market. This project therefore would like to build on such traits, especially in the context of multi-sided groups.The aims and objectives of the researchThis project aims at dveloping computational analysis, simulation and intervention of interaction dynamics in digital ecosystems, seeking to answer the following key problems;1. How to assess the inherant role each individual/organisation. complay plays in a system-levale interpretable way.2. Can we make meaningful comparison between milti-sided groups via thier role sturctures and behavioral simiulation.3. To improv the digital ecosystem, what intervention approaches can be proposed based on the analysis above? This will likely make use to the reommendation algorithm.4. What are the scalability limitations of the method above, and can we mitigate them to enable analysis on increasing data sizes (e.g across large population samples or different geographies)?The novelty of the research methodologyThe main novelty in this research is the application of network structure techniques to assess system-level collective behavioral features in digital ecosystems and propose relevant intervention approaches accordingly, particularly relevant with recmmendation algorithms.The potential impact, applications and benefitsIntroducing complex network structures to digital ecosystems can provide an intuitive system-level evaluation of extensive user interaction dynamics. Appropriate intervention methods may be explored to potentially improve digital regilation and policy.How the rsearch relates to the remitThis research is closely related with the EPSRC remit in both its methodology and goal. Complex network modelling as the primary methodology strongly relies on mathematical sciences. The goal is to improve the digital ecosystem, lying in both Digital Economy and ICT (Inofrmation and Communication Technologies).External Partner - StateUP (https://stateup.co/nebula/) - StateUp is a start up company which stands at the crossroads of industry and academia, working on procurement technologies for the publc sector. It will provide me with both internship opportunities and datasets. In details, I can help build Nebula, a repository curated by stateUp that can be used to make recommendations to public bodies seeeking to resolve societal problems at various scales. During this process, I will also have exclusive access to acurated dataser of company profiles.