Computer testing of candidates for job position based on neural networks

Lukyanov Leonid Mikhailovich

The purpose of the study is to develop an approach to computer testing of candidates for a position in an organization using artificial neural networks technology. The problem of the increasing complexity of finding the most effective candidates for a position at the present time, due to the excess of demand over supply in the labor market, is outlined. The existing neural network approaches to assessing candidates for a position have been analyzed. An approach to computer testing of candidates for a position in an organization is described, which uses the technology of artificial neural networks for direct assessment of test results, which takes into account not only the correctness of answers to questions, but also the nature of the candidate passing the test. The set of input data on passing the test by the candidate is described, which is taken into account by the developed system. An algorithm for processing test results is described. It also describes a set of outputs reflecting the degree to which a candidate meets the requirements of the position for which he is applying.

Keywords:testing, candidate for a position, performance assessment, neural networks