Applications of social data science to environmental communication and activism on social media

我提出的博士研究是开发一种新的社会数据科学模型和框架，目的是研究和解释跨文化背景下在线气候变化交流的独特挑战动态。该模型将利用深度学习和网络分析范式来识别和描述不同地理区域的气候变化话语，并使用定性分析来控制不同背景下独特的文化、政治和社会特征。最终目标是利用这些信息，在给定一个地区独特的环境属性的情况下，预测在线传播气候变化内容的最有效策略。出于多样性和普遍性的目的，重点国家包括那些在制定气候变化内容和立法行动方面最活跃的国家，以及那些受气候变化后果影响最直接的国家。到目前为止，大多数研究气候变化在线传播的计算方法都局限于对话语的两极分化和隔离的描述性分析；我们的新范式试图通过开发一种工具和一般研究框架来超越这一点，这将使问题以及其他类似的复杂和依赖于上下文的在线社会现象能够得到更好的解释和预测分析。这项研究的影响是双重的。首先，它将通过为其他研究人员提供一个开源模型来使用和扩展，从而为数据科学研究的技术工具箱做出贡献。这项工作将开创一种先进的跨学科方法，以伦理和文化包容的计算社会科学。它还将证明社会数据科学是社会科学和数据科学的真正共生组合，而不是为了适应数据科学方法而扭曲社会科学问题。

My proposed doctoral research is the development of a novel social data science model and framework for the purpose of studying and explaining the uniquely challenging dynamics of online climate change communication across cultural contexts. The model will leverage deep learning and network analysis paradigms to identify and characterise climate change discourse in various geographic regions, using qualitative analysis to control for unique cultural, political,and social features of different contexts. The ultimate objective is to use this information to predict the most effective strategies for disseminating climate change content online given a region's unique contextual attributes. For the purposes of diversity and generalisability, the countries of focus include those that have been most active in generating climate change content and legislative action, along with those that have been most directly affected by theconsequences of climate change. Thus far, most computational approaches to studying climate change communication online have been limited to descriptive analysis of the polarisation and segregation of the discourse; our new paradigm seeks to go beyond this by developing a tool and general research framework that will enable greater explanatory and predictive analysis of the issue, as well as other similarly sophisticated and context-dependent online socialphenomena.The impact of this research is two-fold. First, it will contribute to the technical toolbox of data science research by providing an open source model for other researchers to use and augment. This work will pioneer an advanced interdisciplinary approach to ethical and culturally-inclusive computational social science. It will also prove social data science as a truly symbiotic combination of social and data science and not a distortion of social science questions to fit data science methods.