Smart4Job: A Big Data Framework for Intelligent Job Offers Broadcasting Using Time Series Forecasting and Semantic Classification

Recently, dedicated web portals and social networks for the automatization of the recruitment processes, have emerged with the expansion of the Internet, leading to a wide use of optimized algorithms. To that aim, a lot of job boards websites have been created, for disseminating and sharing at the best the job offers. Choosing the relevant job board for the broadcasting of a given job offer can be sometimes difficult for the recruiters, since they always seek to attract the best candidates, in a short period of time. Moreover, some job boards have different business categories, which can make the selection very difficult. To deal with these problems, we propose in this paper, Smart4Job a new job boards recommendation system, which proposes the adequate job boards for the dissemination of a new job offer. Our system is based on a hybrid representation on a big data platform, which includes both a domain knowledge analysis and a temporal prediction model. The semantic classification of job boards requires a textual analysis using a controlled vocabulary. The time series analysis module is used to predict the best job board for a given offer, using the history of the clicks. The answers of these modules are combined during the decision making process. The proposed system has been evaluated on real data, and preliminary results seem very promising.