Application for Professionals Search Based on Machine Learning Algorithms

This paper is dedicated to the process of a predictive mathematical model engineering. This model is aimed on the prediction of an applicant correspondence to the specific vacancy and successful accomplishment of job requirements by concrete candidate. In this work, different approaches to the machine learning techniques implementation to e-recruitment systems and candidate ranking problem are provided. Moreover, the basic iterative process of supervised predictive model engineering is presented. Each step of this process is represented by the theoretical basis for applied machine learning algorithms and data mining techniques. This work is performed according to the PROFI.RU company task on the development of a predictive service. This company is one of the biggest service marketplaces in Russian internet market segment, which provides an opportunity to search for a specialist to do a specific job. In accordance with given task, different techniques used in this work are based on the production data provided by the company. The main objective of this work is the development of an application, which provides the functionality on prediction of successful job order execution by a concrete specialist of the PROFI.RU resource with predetermined accuracy. The developed service has an interaction API in order to cooperate, and integrate with internal company production systems, and be deployed in the resource VPN. Key words: e-recruitment, recruitment systems, machine learning, data analysis.