I-Corps: Honest Signals-Machine Learning for Job Candidate Assessment

I-Corps：用于求职者评估的诚实信号-机器学习

• 批准号: 1511655
• 负责人: David Cox
• 金额: $ 5万
• 依托单位: Harvard University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2015
• 资助国家: 美国
• 起止时间: 2015-01-01 至 2015-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1511655&HistoricalAwards=false
• 关键词: Corps; Honest; Signals; Machine; Learning

A recent study on hiring indicated that interviewers felt that 80 to 85 percent of the job candidates they interviewed were poorly prepared. This has led to a remarkable rise in career coaching and employment websites. At the present time, the American unemployment rate is 6.2%, representing 9.7 million unemployed persons currently seeking work. Extending beyond this figure, the underemployment rate, or the overall rate of those seeking full-time work, is estimated to be as high as 15.1% for the same period. The vast majority of these job seekers lacks experience in branding themselves successfully and does not possess an extensive professional network from which to learn about job opportunities; this means that companies could easily be missing out on large numbers of qualified candidates. The proposed technology provides digital tools to help candidates hone their interviewing skills and professionalism for no charge, while simultaneously providing companies with an inexpensive platform to apply video analytics to the task of pre-screening candidates. This will allow companies to gain the high-value information normally only obtainable at high cost during a live interview for a much broader set of candidates.This I-Corps team envisions a free service where job candidates are able to try the proposed concept "Honest Signals" video analytics features before subscribing and gaining access to a network of potential employers. Companies who use the Honest Signals service for employee screening will have access both to a network of users who have asked to have their videos shared with employers and to custom analytics tools which will allow them to screen candidates based on their own unique needs, thus streamlining human resources operations for challenging hiring scenarios such as low-end/unskilled labor, retail associates, and overseas labor. The proposed work will bring fundamentally new kinds of information to job candidate assessment and will help pioneer automated methods to exploit these new data sources. The use of these vastly richer forms of information has the potential to transform the practice of hiring through the use of innovative techniques drawn from social psychology.

最近一项关于招聘的研究表明，面试官认为他们面试的求职者中有80%到85%的人准备不足。这导致了职业指导和就业网站的显著增长。 目前，美国失业率为6.2%，有970万失业者正在寻找工作。除此之外，同期的就业不足率，即寻求全职工作的总比率，估计高达15.1%。 这些求职者中的绝大多数缺乏成功塑造自己品牌的经验，也没有广泛的专业网络来了解工作机会;这意味着公司很容易错过大量合格的候选人。这项技术提供了数字化工具，帮助候选人免费磨练他们的面试技巧和专业精神，同时为公司提供了一个廉价的平台，将视频分析应用于预选候选人的任务。这将使公司能够获得高价值的信息，通常只有在现场面试中才能获得高成本的更广泛的候选人。这个I-Corps团队设想了一个免费的服务，求职者可以尝试拟议的概念“诚实的信号”视频分析功能，然后订阅并访问潜在雇主的网络。使用Honest Signals服务进行员工筛选的公司将可以访问要求与雇主共享视频的用户网络，以及自定义分析工具，这些工具将允许他们根据自己的独特需求筛选候选人，从而简化人力资源运营，以应对具有挑战性的招聘场景，如低端/非熟练劳动力，零售员工和海外劳动力。拟议的工作将为求职者评估带来全新的信息，并将有助于开拓利用这些新数据源的自动化方法。使用这些丰富得多的信息形式，有可能通过使用来自社会心理学的创新技术来改变招聘做法。