SBIR Phase I: Virtual Interview Trait Estimation Combining Speech and Touch

这一小型企业创新研究(SBIR)第一阶段项目的更广泛影响/商业潜力在于，通过更快、更准确地评估求职者、员工和学生的人际关系特征(“软技能”)，提高了劳动力市场的效率。该项目开发了准确的自动筛选面试方法，可以为雇主和求职者节省时间、精力和金钱-为每年对求职者进行数亿次筛选面试的美国公司和公共机构提供更高效的候选人选择流程。这个项目的重点是对态度、精力和社会沟通的绩效评估，特别是因为这些素质是在就业环境中挑选合作角色所必需的。该项目的应用研究将完善(并可能协调)在员工选择中有价值的特质的行为定义，因为它们被认为是在工作环境和职业培训中取得成功的预测。由此产生的精细化的测量工具可能会推动工业心理学等应用领域的进步，并为情绪、社会知觉、个性和精神障碍的神经学相关性的基础研究提供更精确的变量。该小型企业创新研究(SBIR)第一阶段项目设计、开发并部署了一个进行自动虚拟筛选面试(VSI)的原型应用程序。VSI应用程序是一种实验性工具，可以探测语音和触摸反应中与工作相关的特征。近几十年来，算法已经通过面部表情、文本内容、语调、语速、手势、姿势和其他肢体语言来估计人们的情感、情绪和情绪。其中，语音韵律、语言内容和面部表情已被证明是情绪效价(态度)和精神运动激活(能量水平)的有力指标。VSI系统研究将确定在需要团队合作或社交技能的活动和职业中支持快速学习和成功表现的可测量特征。VSI应用程序收集移动设备上的性能数据(语音和屏幕触摸轨迹)，以研究语音和运动功能的新颖组合，这些功能可以加速和改进特征测量。该项目将产生关于语言和动作之间的关系如何与工作组和社会环境中的人的专家评估相匹配的新知识。该项目开发的测量技术还可以应用于教育和工业，以提高教学和培训的效率和效果。

The broader impact/commercial potential of this Small Business Innovation Research (SBIR) Phase I project is in improved efficiency in labor markets by enabling faster, more accurate assessments of the interpersonal traits ("soft skills") of job candidates, employees, and students. The project develops methods for accurate automated screening interviews that can save time, effort, and money for both employers and job seekers - delivering a more efficient candidate selection process for U.S. companies and public agencies that conduct hundreds of millions of screening interviews with job candidates each year. This project is focused on performance assessments of attitude, energy and social communication, specifically as these qualities are needed for selection of people for cooperative roles in employment settings. The project's applied research will refine (and may reconceive) the behavioral definitions of traits that are valued in employee selection because they are held to be predictive of success in work settings and in occupational training. Resultant refined measurement tools may enable advances in applied fields like industrial psychology and provide more precise variables for use in basic studies of the neurological correlates of emotion, social perception, personality, and mental disorders. This Small Business Innovation Research (SBIR) Phase I project designs, develops and fields a prototype app that conducts an automated Virtual Screening Interview (VSI). The VSI app is an experimental instrument that probes job-relevant traits in speech and touch responses. In recent decades, algorithms have estimated people's affect, sentiment, and emotions as observed in facial expression, text content, tone of voice, speech rate, gestures, posture and other body language. Among these, speech prosody, linguistic content, and facial expression have proven to be strong indicators of emotional valence (attitude) and psychomotor activation (energy level). VSI system research will identify measurable traits that support rapid learning and successful performance in activities and occupations that require teamwork or social skill. The VSI app collects performance data (voice and screen-touch tracks) on mobile devices to investigate novel combinations of features from speech and motion that accelerate and improve trait measurement. The project will generate new knowledge about how the relationships between speech and movement can match expert evaluations of people in work groups and in social settings. Measurement technologies developed in the project may also be applied in education and industry to improve the efficiency and effectiveness of instruction and training.