A revolutionary AI Driven Workplace Anti-Bully System, Predicting, Or Reporting Incidents Through Machine Learning

革命性的人工智能驱动的工作场所反欺凌系统，通过机器学习预测或报告事件

• 批准号: 87744
• 金额: $ 12.73万
• 依托单位: DARK MATTER DATA LIMITED
• 依托单位国家: 英国
• 项目类别: Collaborative R&D
• 财政年份: 2020
• 资助国家: 英国
• 起止时间: 2020 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=87744
• 关键词: revolutionary; AI; Driven; Workplace; Anti

Founded by Matthew Barnett and Andrew Church, DarkMatter is a UK-based SME that intends to address the current challenges in delivering anti-bully and harassment strategies in the increasing environment of homeworking. Offering backgrounds in IT consultancy, Big Data NLP and ML, and Wellbeing Programmes, the founders of this project propose to develop a machine learning product, which intends to reduce bullying and harassment, and utilise natural language processing. This project could improve employee wellbeing on a wide scale and support organisations' responsibility to address issues of inequality and discrimination. DarkMatter anticipates a Year-5 post-project revenue of £5.1M.**Scope/Summary**Current approaches to anti-bully and harassment prevention are, in many cases, not digitalised and not equipped to address employee needs in the increased landscape of homeworking. DarkMatter's software will address this issue by using NLP to analyse employees' behaviour and flag when safeguarding and/or wellbeing is being impeded.**Innovation and EDI Considerations**DarkMatter's remote worker/learner bullying and harassment software offers an ethically-focused listening application to standard PCs, which harvests and securely transmits written/spoken data to its advanced analytics platform. Due to the sensitivity of the data sourced, this is hosted on secure on-premise infrastructure. The development of NLP, Neural Networks, and Machine Learning models, at a depth that was previously unavailable, has ensured the development of a solution that vectors keywords and phrases. Employee sentiments and emotions can be classified against a bullying and harassment behavioural baseline, and subsequently providing an automated alerting where a concern is flagged. In addition, team level behavioural patterns may be mapped and analysed over time to identify abnormalities relating to the inclusion and consistency of communication with specific individuals.There are no identified potential or inadvertent racial/ethnic/gender biases within our proposal that could negatively affect any one gender or ethnic group. This will continue to be monitored throughout the project.**Market Opportunity and Sustainability**Work-related bullying and harassment are reaching epidemic levels. There is currently no solution to holistically monitor and safeguard remote working teams against either. DarkMatter's innovation will be first to market.DarkMatter's innovation will help to maintain the heightened levels of remote working and, consequently, reduce the need for business travel and its associated carbon footprint.The global and UK annual addressable/serviceable markets are £818.28M/£818.28M and £145M/£50M, respectively (CAGR: 7.1% 2019-2027; The Insight Partners).

DarkMatter 由 Matthew Barnett 和 Andrew Church 创立，是一家总部位于英国的中小企业，旨在应对当前在家办公日益增多的环境中提供反欺凌和骚扰策略的挑战。该项目的创始人拥有 IT 咨询、大数据 NLP 和 ML 以及福利计划的背景，提议开发一种机器学习产品，旨在减少欺凌和骚扰，并利用自然语言处理。该项目可以广泛改善员工福祉，并支持组织解决不平等和歧视问题的责任。 DarkMatter 预计项目后第 5 年的收入为 510 万英镑。**范围/摘要**当前的反欺凌和骚扰预防方法在许多情况下都没有数字化，也无法满足员工在家办公日益增多的需求。 DarkMatter 的软件将通过使用 NLP 分析员工的行为并在保护和/或福祉受到阻碍时进行标记来解决此问题。**创新和 EDI 考虑因素**DarkMatter 的远程员工/学习者欺凌和骚扰软件为标准 PC 提供注重道德的监听应用程序，该应用程序收集书面/口头数据并将其安全传输到其高级分析平台。由于来源数据的敏感性，该数据托管在安全的本地基础设施上。 NLP、神经网络和机器学习模型的发展达到了前所未有的深度，确保了矢量关键词和短语解决方案的发展。可以根据欺凌和骚扰行为基线对员工的情绪和情绪进行分类，然后在标记问题时提供自动警报。此外，随着时间的推移，可以对团队层面的行为模式进行映射和分析，以识别与特定个人沟通的包容性和一致性相关的异常情况。我们的提案中没有发现可能对任何一种性别或族裔群体产生负面影响的潜在或无意的种族/民族/性别偏见。这将在整个项目中继续受到监控。**市场机会和可持续性**与工作相关的欺凌和骚扰已达到流行程度。目前还没有解决方案可以全面监控和保护远程工作团队免受这两种情况的影响。 DarkMatter 的创新将率先进入市场。DarkMatter 的创新将有助于保持远程工作的高水平，从而减少商务旅行的需求及其相关的碳足迹。全球和英国的年度可寻址/可服务市场分别为 81828 万英镑/81828 万英镑和 1.45 亿英镑/5000 万英镑（复合年增长率：2019-2027 年复合年增长率：7.1%；Insight Partners）。