E-AVI improves video interview scoring with transparent evidence support
E-AVI: Evidence-Grounded Multimodal Assessment for Automated Video Interviews
Artificial Intelligence
Summary
Scoring video interviews automatically is hard because verbal answers, tone, and facial expressions all matter, but numbers alone don’t explain the reasons behind scores. The authors created E-AVI, a system that highlights specific moments from the video as evidence for the scores, helping people understand why a certain rating was given. This evidence also supports providing feedback in natural language and answering follow-up questions. Their tests show that E-AVI scores videos more accurately than earlier methods and provides useful explanations.
What this means in practice
- •For recruitment platform developers: Integrate transparent scoring with explainable evidence to improve candidate evaluation accuracy and user trust in automated video interviews.$Commercial implications: Creates a new product feature that enhances automated interview platforms by combining scoring with inspectable evidence to support hiring decisions.
- •For hospitality training managers: Use evidence-grounded scoring to provide actionable feedback and follow-up questions from employee video assessments.
Authors
Haoshen Wang, Dongbo Che, Zeyi Xie, Yuanjie Du, Shicheng Hua, Xingyu Wang
Abstract
Automated video interview assessment integrates verbal content, acoustic delivery, and visual behavior, yet numerical predictions alone provide limited inspectable support. We present E-AVI, an evidence-grounded framework that extracts timestamped multimodal evidence and integrates dimension-conditioned evidence attention with source-level embeddings for scoring. A shared evidence pool further supports natural-language feedback and follow-up question answering. On RecruitView and a private hospitality dataset, E-AVI consistently outperforms fine-tuned multimodal baselines in rank correlation. Ablation, evidence-deletion, bootstrap, human-audit, and QA analyses characterize the predictive contribution, grounding, and practical utility of the evidence pathway. Together, these results demonstrate that our proposed E-AVI framework improves predictive performance while providing inspectable support for assessment, feedback, and interactive analysis.