Collaborative Research: FW-HTF-R: The future of news work: Human-technology collaboration for journalistic research and narrative discovery

这个项目探索了新闻工作和计算工具未来的共同发展。记者和新闻工作值得研究，因为即使面对众多挑战，新闻机构仍然是维护社区安全和健康的强大力量，它们提供有关事件和趋势的信息，帮助人们对自己的生活做出更好的决定。新闻业作为一个以信息为基础的行业和职业，一直受到新技术的影响，最近的影响是互联网的兴起和广告收入的损失，以及新闻价值观和商业利益之间长期存在的紧张关系。这些压力对新闻流程和产品造成了影响。一些重要的故事没有被讲述：它们需要太多的时间来发现、研究和精心编成叙事，而且时间越来越短缺。该项目力求通过实地调查和以用户为中心的互动工具开发相结合的方式来解决这些问题，以支持记者的研究和叙事发现，即找到数据所讲述的故事以及将该故事传达给读者的有趣和相关的方式。该项目将通过实地研究和以用户为中心的互动工具的开发，促进技术和新闻工作的知识，以支持记者的研究和叙事发现。由此产生的技术支持的工作流程将在自由撰稿人的实验中进行评估，并通过部署在新闻编辑室进行评估。拟议的项目由几个相互交织的组成部分组成，每个组成部分都有特定的重点，但旨在相互支持并导致研究趋同：(1)侧重于更好地了解工作人员、他们的工作，特别是他们目前与新闻编辑室技术的关系；(2)建立和评估支持新闻研究的方法，特别是利用深度学习技术在不同类型的大型文件集中找到相关信息，用于实体提取和共指解析(AllenNLP)、语义搜索(BERT)和常识推理(COMET)；以及(3)建立和评估支持叙事发现的方法。所产生的工具将在实地试验中进行评估，并通过它们对实地研究中审查的工作的影响进行评估。这两套工具稍后都将通过添加处理音频数据的功能进行扩充。将扩大评估范围，以考虑在人类记者和计算工具之间分配任务的其他方法。贯穿整个项目的将是评估、综合、传播结果和工具、规划和项目管理方面的工作。这一奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

This project explores the future co-evolution of news work and computational tools. Journalists and news work are important to study because, even in the face of numerous challenges, news organizations remain a powerful force for keeping communities safe and healthy by providing information about events and trends to help people make better decisions about their lives. Journalism, as an information-based industry and profession, has always been impacted by new technologies, most recently by the rise of the Internet and loss of advertising revenue, as well as the long-standing tension between journalistic values and business interests. These pressures take a toll on journalistic processes and products. Some important stories are not getting told: they take too much time to discover, research, and craft into narratives, and time is increasingly in short supply. The project seeks to address these problems through a combination of fieldwork and user-centered development of interactive tools for supporting journalists in research and narrative discovery, i.e., finding the story that the data tell and interesting and relevant ways to convey that story to a reader. This project will advance knowledge about technology and news work through field studies and user-centered development of interactive tools for supporting journalists in research and narrative discovery. The resulting technology-supported work processes will be evaluated in experiments with freelance journalists and through deployment in newsrooms. The proposed project is composed of several interwoven components, each with a specific focus, yet designed to be mutually supportive and to lead to research convergence: (1) focus on better understanding workers, their work and especially their current relationship to technology in newsrooms; (2) build and assess ways to support journalistic research, specifically finding relevant information in large document collections of different kinds, using deep learning techniques for entity extraction and coreference resolution (AllenNLP), semantic search (BERT) and common sense reasoning (COMET); and (3) build and assess ways to support narrative discovery. Resulting tools will be evaluated in field experiments and through their impacts on work examined in field studies. Both sets of tools will later be augmented by adding the ability to work on audio data. Assessment will be extended to consider alternative ways to divide tasks between human journalists and the computational tools. Running throughout the project will be work on evaluation, synthesis, dissemination of findings and tools, planning and project management.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.