Analyzing and increasing user engagement in online forums

该项目是多伦多大学马尔巴赫教授的研究团队与VerticalScope公司之间的合作。VerticalScope成立于1999年，是一家总部位于加拿大多伦多的私人控股公司，专门从事在线论坛的收购和开发。该项目的目标是开发算法，使VerticalScope能够分析和提高在线论坛的用户参与度，以及分析VerticalScope有兴趣收购的在线论坛的用户参与度/价值。为了实现这一目标，该项目将设计和实现基于多伦多大学Peter马尔巴赫教授研究团队开发的数学模型和结果的新算法。模型开发的方法是一种博弈论方法，其中假设网络中的每个代理/节点决定动作（例如建立哪些链接，加入哪些社区以及在每个社区中发布/阅读哪些内容），以最大化自己的效用。然后，我们使用的纳什均衡的概念来描述所产生的网络结构。行业合作伙伴VerticalScope将在其在线论坛上提供对真实数据的访问，这将使我们能够测试/验证我们的模型和建模假设，以及由此产生的算法。由此产生的算法将在VerticalScope的数据科学研发团队的指导和监督下在VerticalScope的平台上实施和测试。

The project is a collaboration between the research team of Prof. Marbach at the University of Toronto and the company VerticalScope. Founded in 1999, VerticalScope is a privately held corporation headquartered in Toronto, Canada, that specializes in the acquisition and development of online forums. The goal of the project is to develop algorithms that allow VerticalScope to analyze and increase the user engagement in their online forums, as well as to analyze the user engagement/value of online forums that VerticalScope is interested in acquiring. To achieve this, the project will design and implement novel algorithms that are based on the mathematical models and results developed by the research team of Prof. Peter Marbach at the University of Toronto. The methodology for the development of models is a game-theoretic approach where it is assumed that each agent/node in the network decides on actions (such as which links to establish, which communities to join, and which content to post/read in each community) in order to maximize their own utility. We then use the concept of a Nash equilibrium to characterize the resulting network structure. The industry partner VerticalScope will provide access to real-life data on their online forums that will allow us to test/validate our models and modeling assumptions, as well as the resulting algorithms. The resulting algorithms will be implemented and tested on VerticalScope's platform under the guidance and supervision of VerticalScope's Data Science R&D team.