The first AI Smart Interviewer in the world, developed by Sapia.ai, will submit four ground-breaking research reports at the Society for Industrial and Organizational Psychology’s (SIOP) annual conference in April 2023.
The research investigations on the application of artificial intelligence for grading text chat-based structured interviews in the applicant selection process are being conducted by the Sapia Labs team.
The research focuses more specifically on how improvements in natural language processing (NLP) can be used to create an interview process that reduces bias, enhances the candidate experience, and furthers the fairness and ethical use of AI with cutting-edge methods for detecting plagiarism and outlier answers.
The following papers have been approved for presentation:
- Reducing Bias and Improving the Experience for Candidates with Disabilities through Disability Hiring
- Finding and reducing racial bias in structured interview answers
- Detection and Frequency of Plagiarism in Online Text-Based Structured Interviews
- Outlier Answer Detection with Question-Awareness for Fairer AI Interview Scoring
The first two articles show how Sapia’s asynchronous, untimed, text-based delivery method and usage of feature scores obtained from AI can aid to lessen bias based on ethnicity and disability status.
The last two studies investigate two sorts of adversarial inputs, including novel methods for identifying and highlighting copied material and unusual, irrelevant answers to questions. These publications all have significant ramifications for enhancing the ethical application of AI.
The latest study, according to Sapia Chief Data Scientist Dr. Buddhi Jayatilleke, would further establish Sapia as a pioneer in moral AI for hiring.
“While the world has been captivated by the power of ChatGPT, Sapia.ai has made significant advances in responsible and ethical use of language based AI in recruitment in the past several years. As foundational large language models like GPT3 become a commodity, Sapia.ai is well placed to extend those advances to chat based interviews with its unique conversational dataset and innovations in responsible AI.”
Dr. Buddhi Jayatilleke, Chief Data Scientist, Sapia.ai
“The research we are presenting in this year’s SIOP conference is a real testament to Sapia’s commitment to world-class research and thought leadership in this space.”
The reports will be presented in Boston, USA, during the 38th Annual SIOP Annual Conference.
The Sapia research is being presented at the Annual SIOP conference for the second time in a row. The Sapia Labs team published a study in 2022 titled Mitigating Gender Bias in Structured Interview Responses that demonstrated how linguistic data from structured interviews may be utilized to grade applicants while minimizing gender biases.
The market-leading Smart Interviewer, a text-based chat interview platform that uses Natural Language Processing (NLP) to analyse candidate responses and create skill and personality profiles for the purposes of accurate, effective hiring, will be advanced by Sapia’s ongoing research and development initiatives.