Systems and Methods for Automating Validation and Quantification of Interview Question Responses

Abstract

In an illustrative embodiment, systems and methods for automating candidate video assessments include receiving a submission from a candidate for an available position including baseline response video segments and question response video segments. The system can determine, from detected nonverbal features within the baseline response video segments, nonverbal baseline scores. For each of the interview questions, candidate response attributes can be detected including a response direction, a response speed, and nonverbal features. A nonverbal reaction score is calculated from the detected nonverbal features and the baseline scores. A response score can be calculated from the response direction and response speed, and a trustworthiness score is determined based on a correspondence between the response score and the nonverbal reaction score. A next interview question can be determined in real-time from a benchmarked version of the response score. Overall scores reflecting candidate trustworthiness can be presented within a user interface screen.

What is claimed is: 
     
         1 . A system for automatically conducting an objective evaluation of nonverbal responses made by a candidate for an available position, the system comprising:
 processing circuitry; and   a non-transitory computer readable memory coupled to the processing circuitry, the non-transitory computer readable memory storing machine-executable instructions, wherein the machine-executable instructions, when executed on the processing circuitry, cause the processing circuitry to
 receive, from a remote computing device of a first party via a network, a video submission from the candidate for the available position, wherein
 the video submission includes a plurality of baseline response video segments and a plurality of question response video segments, and 
 each of the plurality question response video segments responds to one of a plurality of interview questions associated with the available position, 
 
 determine, based on one or more nonverbal features detected within the one or more baseline response video segments, at least one baseline non-verbal reaction score for the candidate, 
 for each question of the plurality of interview questions,
 detect candidate response attributes from the question response video segment for the respective interview question, wherein the candidate response attributes include one or more prosodic features and/or one or more facial expression features, and 
 calculate, from the candidate response attributes and based on the at least one baseline nonverbal expression score, at least one nonverbal reaction score, and 
 
 prepare, for review by a second party, candidate interview results comprising at least one candidate score reflecting the trustworthiness of the candidate responses to the plurality of interview questions based on the nonverbal reaction score for each question of the plurality of interview questions.