Understanding Social Media Monitoring

人们普遍认为，社交媒体越来越多地渗透到我们的日常生活中。社交媒体数据也是如此：个性化的在线广告（如Facebook上的广告）、越来越依赖数据并由数据驱动的新闻业，以及Klout和PeerIndex等衡量我们数字声誉的系统的兴起，只是当前使用社交媒体数据的一些例子。最近一些备受瞩目的案例引发了公众对社交媒体数据一旦被分享会发生什么的兴趣，例如美国大学生的基督教父母在他与同性恋合唱团的联系出现在他的Facebook时间轴上时意识到他的同性恋身份（Wolford 2012）。除了Facebook等社交媒体大型平台对数据的分析和操纵外，地球仪各地还出现了监测、解释和分析社交媒体数据的中介公司，提供广泛的社交媒体“情报”服务，包括监测受众和“影响者”的社交媒体活动，以衡量和回应公众舆论。这种分析涉及特定方法的部署，例如情感分析，意见挖掘，社交网络分析，机器学习和自然语言处理。这些社交媒体监控方法，就像所有类型的方法一样，“由它们所处的社交世界塑造，[并]反过来帮助塑造社交世界”（Law 2010：2）。鉴于这些现象-社交媒体数据分析的传播，公众对社交媒体数据发生的事情的关注的增长，以及方法在塑造社交世界中所扮演的角色-对社交媒体监测（SMM）进行实证研究至关重要，以询问有关社交媒体数据如何收集和构建的问题，它们可以被认为是什么意思，它们的部署目的是什么，谁可以访问它们，以及它们的分析如何受到监管（boyd和Crawford 2011）。我的研究通过关注中介SMM行业来实现这一点。它从多个角度探讨SMM，以描绘一幅比迄今为止SMM行业及其社会和经济意义更完整的图景。这些包括：>在这些行业工作的人的观点（在社交媒体监控工作的人如何思考和处理/围绕隐私和监控等核心问题？）; >用户的观点（社交媒体用户如何看待对其社交媒体活动的监控？）; >监管视角（如何以及应该如何管理社交媒体监控行业？）; >以及数据本身的视角（社交媒体数据会发生什么？如何使用它？有什么影响，如果有的话？）。通过对公司网站和其他相关文件的文本分析、与社交媒体用户、社交媒体监测专业人员和主要监管参与者的访谈和焦点小组讨论，该研究将作为题为《社交媒体监测：社交媒体数据分析和社会生活的生产》的专著出版。在拟议的项目范围内，将通过一系列访问、一次国际会议和一份期刊特刊，启动关于社会媒体数据分析作为媒体研究调查对象的重要性的国际对话。该研究还具有更广泛的，超越媒体研究的变革潜力，因为它所采用的批判性方法，超越并平衡了社交媒体民主可能性的庆祝账户，与受社交媒体无处不在影响的其他学科相关。由于社交媒体监测对非学术受众也很重要，因此研究结果将通过讲座和在线传播在学术界之外进行讨论，面向两个关键受众：社交媒体监测行业本身和社交媒体用户。

It is widely accepted that social media permeate more and more of our daily lives. So do social media data: personalized online advertisements such as those seen on Facebook, journalism which is increasingly dependent on and driven by data, and the rise of systems which measure our digital reputations, such as Klout and PeerIndex, are just some examples of current uses of social media data. A number of recent high profile cases have sparked public interest in what happens to social media data once it has been shared, such as the US college student whose Christian parents became aware of his homosexuality when his association with a gay choir appeared on his Facebook timeline (Wolford 2012). Alongside the analysis and manipulation of data by social media mega-platforms like Facebook, intermediary companies which monitor, interpret and analyse social media data are emerging across the globe, offering a broad range of social media 'intelligence' services, including monitoring the social media activities of audiences and 'influencers' in order to gauge and respond to public opinion. Such analysis involves the deployment of specific methods, such as sentiment analysis, opinion mining, social network analysis, machine learning and natural language processing. These social media monitoring methods, like methods of all kinds, 'are shaped by the social world in which they are located [and] in turn help to shape that social world' (Law 2010: 2). Given these phenomena - the spread of social media data analysis, the growth of public concern about what happens to social media data, and the role played by methods in the shaping of the social world - it is vital that empirical research into social media monitoring (SMM) is carried out, to ask questions about how social media data are gathered and constructed, what they can be taken to mean, to what ends they are deployed, who gets access to them and how their analysis is regulated (boyd and Crawford 2011). My research does this through a focus on the intermediary SMM industries. It approaches SMM from a number of perspectives, in order to paint a more complete picture than exists to date of the SMM industries and their social and economic significance. These include:> the perspectives of people working in these industries (How do people working in social media monitoring think about and work with/around core issues such as privacy and surveillance?); > users' perspectives (What do social media users think about the monitoring of their social media activity?); > a regulation perspective (How could and should the social media monitoring industries be governed?); > and the perspective of the data itself (What happens to social media data? How does it get used? What impact, if any, does it have?). Drawing together textual analysis of company websites and other related documents, interviews and focus group discussions with social media users, social media monitoring professionals and key regulation actors, the research will be published as a monograph entitled Social Media Monitoring: social media data analysis and the production of social life. Within the proposed project, international dialogues about the importance of social media data analysis as an object of investigation for media studies will be initiated, through a series of visits, an international conference and a special issue of a journal. The research also has broader, transformative potential beyond media studies, because the critical approach that it adopts, which moves beyond and counterbalances celebratory accounts of the democratic possibilities of social media, is relevant to other disciplines impacted by the ubiquity of social media. As social media monitoring also matters to non-academic audiences, the research findings will be discussed outside of academia, through talks and online dissemination, to two key audiences: the social media monitoring industries themselves and social media users.