Generating An Optimal Interview Question Plan Using A Knowledge Graph And Integer Linear Programming

Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 1996–2005 June 6–11, 2021. ©2021 Association for Computational Linguistics 1996 Generating An Optimal Interview Question Plan Using A Knowledge Graph And Integer Linear Programming Soham Datta, Prabir Mallick, Sangameshwar Patil d.soham, mallick.prabir, sangameshwar.patil Indrajit Bhattacharya, Girish Palshikar b.indrajit, gk.palshikar, @tcs.com TCS Research Abstract Given the diversity of the candidates and com- plexity of job requirements, and since inter- viewing is an inherently subjective process, it is an important task to ensure consistent, uni- form, efficient and objective interviews that re- sult in high quality recruitment. We propose an interview assistant system to automatically, and in an objective manner, select an opti- mal set of technical questions (from question banks) personalized for a candidate. This set can help a human interviewer to plan for an upcoming interview of that candidate. We for- malize the problem of selecting a set of ques- tions as an integer linear programming prob- lem and use standard solvers to get a solu- tion. We use knowledge graph as background knowledge in this formulation, and derive our objective functions and constraints from it. We use candidate’s resume to personalize the se- lection of questions. We propose an intrin- sic evaluation to compare a set of suggested questions with actually asked questions. We also use expert interviewers to comparatively evaluate our approach with a set of reasonable baselines. 1 Introduction A large multi-national IT company added roughly 70,000 employees in FY2018-19.1 Assuming an average interview time of 30 minutes, 3 interview- ers in each interview, and 4 candidates interviewed for every position, implies approximately 420,000 person-hours were spent in one year just on con- ducting the interviews. Given the diversity of the candidates and complexity of job requirements, and considering that interviewing is an inherently human and subjective process, it is a mammoth task to ensure consistent, uniform, efficient and objective interviews that result in high quality re- 1https://www.tcs.com/content/dam/tcs/investor- relations/financial-statements/2018-19/ar/annual-report- 2018-2019.pdf cruitment. AI and ML technologies are increas- ingly playing important roles in helping improve recruitment quality, e.g., Faliagka et al. (2012), Javed et al. (2015), Palshikar et al. (2017), al- though ethical issues are emerging.2 In this paper, we consider one particular way to assist human in- terviewers in improving the quality of their inter- views and in reducing subjectivity. Before conducting an interview, an interviewer typically studies the candidate’s resume, noting salient points about her education, skills, job his- tory, roles, projects, tasks handled etc. The inter- viewer also notes the apparent strengths and weak- nesses of the candidate, as also the extent to which she matches (and does not match) the job profile for which she will be interviewed. In short, the in- terviewer builds an a priori rough mental profile of the candidate, and prepares a mental plan of how to interview her. Such a plan includes preparing an unordered set of questions that the interviewer would like to ask the candidate. In this …

Content truncated to 500 words