Reel Lives: personal documentaries constructed from digital data.

新的社交媒体导致了个人数字数据的爆炸式增长，其中既包括个人选择的自我表达，也包括其他第三方提供的自我反映。由此产生的数字人格(DP)数据非常复杂，对于许多用户来说，太容易迷失在数字数据的泥潭中。这是不幸的，因为这些自我的数字表现不仅作为一种自我反省的手段对个人来说很有价值，而且作为社区、商业和社会政策资源也具有巨大的信息价值。在这项提案中，我们认为，以有意义的形式表示和编辑个人数字数据的能力对于个人和社区的赋权至关重要，它还可以为政府和企业带来重大好处。我们还认为，这种“有意义的形式”应该包括一种易于理解和编辑的电影叙事结构，并允许将一种数字生活与另一种数字生活进行简单的比较。简而言之，我们建议从现有的数字数据中构建可搜索的、类似于个人电影的纪录片(我们将其称为“Reels”)。要构建Reels，我们必须首先开发捕获在线生活日志、事件和相关数据的方法。要做到这一点，已经存在多种方法，但我们必须确保我们能够获取分布在多个社交网络上的数据，并将其整合到单个配置文件中，以捕获用户在自己的个人时间线中的不同时间点的兴趣和体验、社交网络、行为、个性和情绪状态。然后，这些配置文件可以为我们提供对用户在不同时间点的生活事件和背景的丰富表示和理解，我们可以进一步处理这些事件和背景，以生成视频故事情节和叙事。简单的时间线结构本身不足以创造有意义的叙事，这里的一个关键技术问题是现有的数字材料在多大程度上可以与现有的已知叙事结构相匹配，以便理解事件和图像的分散景观。有了正确的叙事结构，对象、图像和事件就可以栩栩如生。我们的目标是开发系统，可以将可用的个人资料数据转换为有意义的个人纪录片(Reels)，个人可以自己编辑或移交给电影和媒体专业人员，以创造内在的艺术价值。这个项目的关键是每个Reel都是完全可搜索的，因为它是从底层语义数据生成的。这允许直接比较一个人与另一个人的数字生活，或促进收集和比较数字相似的卷轴，从而加强对社区和其中的公民的了解。我们建议用户和利益相关者与企业和社区领导人进行一段重要的接触，以了解此类比较的价值。

New social media has led to an explosion in personal digital data that encompasses both those expressions of self chosen by the individual as well as reflections of self provided by other, third parties. The resulting Digital Personhood (DP) data is complex and for many users it is too easy to become lost in the mire of digital data. This is unfortunate as these digital representations of self are not only valuable to the individual as a means of self-reflection, but also have great information value as a community, business and social policy resource. In this proposal, we argue that the ability to represent and edit personal digital data in a meaningful form is crucial to the empowerment of individuals and communities and that it could also generate significant benefit to government and business. We would also argue that, this 'meaningful form' should include a filmic, narrative structure that is easy to comprehend and edit and that allows for a simply comparison of one digital life with another. In short, we propose building searchable, personal film-like documentaries (that we would term 'Reels') from existing, digital data.To build the Reels, we must first develop the means of capturing online life logs, events and associated data. A variety of means already exist to do this, but we must ensure that we are able to take data that is distributed over multiple social networks and integrate it into a single profile that captures a users interests and experiences, social networks, behaviour, personality and mood states at various points in their own personal timeline. These profiles can then provide us with rich representations and understanding of users' life events and contexts at different points in time, which we can process further to generate video storylines and narrative. A simple time-line structure is insufficient in and of itself to create a meaningful narrative, and a crucial technical issue here regards the extent to which existing digital material can be matched to existing, known narrative structures in order to make sense of a diffuse landscape of events and images. With the right narrative structures in place, the objects, images and events can then be brought to life.Our aim is to develop systems that can convert the available profile data into meaningful personal documentaries (Reels) that individuals can either edit themselves or that can be turned over to film and media professionals to create Reels of intrinsic artistic value. Crucial to this project is the fact that each Reel is completely searchable, because it is generated from underlying semantic data. This allows for the direct comparison of one individual digital life with another or facilitates the collection and comparison of digitally similar Reels that allow enhanced understanding of communities and the citizens within them. We propose a significant period of user and stakeholder engagement with business and community leaders to understand the value of such comparisons.