Social- Semantic Analysis of Social Media Interactions to assess customer satisfaction in transit agencies

本研究的目的是通过开发一个集成和互动的模型，强调社会和语义（及其情感）网络参数之间的相互作用，来推进社交媒体分析学科的发展。其目的是为交通运输机构提供一种手段，根据不同乘客群体在Twitter和Facebook上的互动，了解他们讨论的话题、关心的问题或感兴趣的问题。也就是说：谁在不同的时间谈论什么话题？讨论的主题是否与最近的事件、通讯或交通系统的变化有关？这对于更好地分析客户及其满意度水平是必不可少的，也是理解行为模式和驱动因素的一步。同样重要的是，这对于有效的沟通和推广至关重要，可以帮助减少客户在社交媒体上典型的负面输入。研究结果将用于建立一个自适应仪表板形式的可定制分析门户，供Transit员工使用，以分析社交媒体内容。预计工作人员将从选择要分析的社交媒体渠道或账户开始。然后，系统将允许用户选择几个语义和社会参数来研究。该系统还允许用户选择和定制定期报告来跟踪趋势。在进行此类分析时使用机器学习可以为系统规划者和操作员提供有价值的见解。

The aim of this proposed study is to progress the discipline of social media analytics in transit through developing an integrated and interactive model with emphasis on the interplay between social and semantic (and their sentiment) network parameters. The aim is to provide transit agencies with means to understand the topics of debate or issues of concerns or interest by different groups of riders based on their interactions on Twitter and Facebook. That is: who is talking about what topics at different times? Do the topics discussed relate to recent events, communications, or changes to the transit system? This is essential to better profiling customers and their levels of satisfaction as a step towards understanding behavior patterns and drivers. Equally important, this is essential for effective communication and outreach that can help reduce the typically negative input by customers on social media.The results of the research will be used to build a customizable analytics portal in the form of an adaptive dashboard for use by Transit staff to analyze social media contents. It is expected that a staff member will start by choosing a social media outlet or account to analyze. The system will then allow the user to select several semantic and social parameters to study. The system will also allow the user to select and customize periodical reports to track trends. The use of machine learning in conducting such analysis can yield valuable insights to system planners and operators.