Social Network Analysis in the field of Political Sociology Case: Brexit and Online Social Networks

本研究的目的是：1.调查与2016年欧盟公投相关的在线社交网络对政治讨论和政治极化的影响;2.测试社交媒体是否鼓励网络同质性和回声室的形成（假设1）或提供一个有效的公共领域，支持民主（假设2）。第一项任务将是描述和分析活跃在英国退欧讨论中的社交媒体用户的在线网络结构。随后，我想研究社交媒体对政治舆论创造的影响和力量。我将研究这些平台（如Twitter）如何影响公众舆论的形成，促进或限制有关政治话题的辩论。此外，我想检查是否在线社交网络平台促进政治两极分化，回声室，从而导致网络结构的特点是高度隔离的集群，只有有限的连接，或者这些平台是否促进政治辩论。换句话说，这项研究将审查互联网是否提供了一个充分和有效的在线公共领域，供个人辩论和交换意见。另一种假设是，这些平台提供了一个形成同性恋关系的背景，这反过来又促进了极端政治观点和两极分化的发展。此外，该研究旨在寻找经验证据来测试对过滤泡沫影响投票行为和整个民主社会的担忧是否有道理。我将通过关注2016年英国脱欧公投和英国脱欧过渡期来回答这些问题。到目前为止，英国是第一个退出欧盟的国家。这一政治事件将对英国法律、经济产生重大影响，最重要的是，它影响到许多英国和欧盟公民。此外，一些人担心其他欧盟成员国也会效仿英国，离开欧盟。研究这一特殊事件不仅因为其对公民的影响而重要，了解英国脱欧背后的机制也可能有助于预测或处理未来类似的政治事件。研究成果在这一节中，我列出了我想实证检验的研究问题：社交媒体平台是否促进了同性恋网络的形成？2)社交媒体是否会损害公共领域，或者Twitter和YouTube等社交网络服务是否为人们提供了一个公开和可访问的平台，供人们辩论和交换政治观点？3)在线公共领域信息传播的基本机制和手段是什么？4)通过社交媒体或网络搜索引擎消费新闻是否会增加意识形态隔离，并导致选民较少接触各种政治观点和观点？5)讨论英国脱欧相关问题的更广泛的英国Twitter社区是否分为两个不同的社区，或者是否存在一个综合的社会领域？6)如果存在隔离的群体，这些群体的隔离程度有多强，是否可以观察到回声室？为了实证检验上述研究问题，我打算采用创新的研究方法，包括社会网络分析（SNA）和自然语言处理（NLP）技术。我将使用SNA来调查和可视化在Twitter等社交媒体平台上活跃于英国退欧讨论的个人网络。网络可以用开源软件Gephi可视化。NetworkX是一个基于Python的软件包，我可以用它来分析网络结构。例如，我将衡量一个给定用户对社区的潜在影响以及网络同质性的程度。

The aim of this study is:1. investigating the effects of online social networks related to the EU Referendum in 2016 on political discussions and political polarisation;2. to test whether social media encourage the formation of network homophily and echo chambers (Hypothesis 1) or provide an effective public sphere that supports democracy (Hypothesis 2).The first task will be describing and analysing the online network structure of social media users who are active in Brexit discussions. Subsequently, I would like to investigate the influence and power social media have on political opinion creation. I will examine how such platforms (e.g. Twitter) influence public opinion formation and facilitate or constrain debates about political topics. Moreover, I would like to inspect whether online social network platforms facilitate political polarisation, echo chambers and in turn lead to a network structure characterised by highly segregated clusters with only limited connectivity, or whether such platforms facilitate political debate. In other words, the study will examine whether the Internet provides an adequate and effective online public sphere where individuals debate and exchange views. The alterative hypothesis would be that such platforms provide a context in which homophilous relations are formed, which in turn foster the development of extreme political views and polarisation. Moreover, the research aims to find empirical evidence to test whether concerns about filter bubbles influencing voting behaviour and the democratic society at large are justified. I will answer these questions by focusing on the UK EU Referendum of 2016 and the Brexit transition period. So far, the UK is the first country who withdraws from the European Union. This political event will have significant consequences on UK law, economics and most importantly it affects a lot of UK as well as EU citizens. Moreover, some fear that other EU member states will follow the UK and leave the European Union as well. Studying this particular event is not only important because of its effect on citizens, understanding the mechanisms behind Brexit might also help to forecast or deal with similar political events in the future.Research QuestionsIn this section I list research questions I would like to test empirically:Do social media platforms facilitate the formation of homophilous networks? 2) Do social media harm the public sphere or do social network services such as Twitter and YouTube provide an open and accessible platform for people to debate and exchange political opinions?3) What are the underlying mechanisms and means of information diffusion in an online public sphere?4) Does the consumption of news via social media or web search engines increase ideological segregation and lead to voters who are less exposed to a variety of political views and perspectives?5) Is the broader UK Twitter community discussing Brexit-related issues differentiated into two distinct communities or does an integrated social sphere exist?6) If segregated groups exist, how strong are these groups isolated and can echo chambers be observed?To test the above-mentioned research questions empirically, I intend to employ innovative research methods including Social Network Analysis (SNA) and Natural Language Processing (NLP) techniques. I will use SNA to investigate and visualise networks of individuals active in Brexit discussions on social media platforms such as Twitter. Networks could be visualised with the open source software Gephi. NetworkX is a Python based package which I could use in order to analyse network structures. For instance, I will measure the potential impact of a given user on its community as well as the degree of network homophily.