The public agency SERVEF has developed a pilot project with Big Content technologies and semantics to reduce unemployment.
The SERVEF is an agency of the Valencian government whose mission is to improve the employability of citizens. The agency carries out labor market surveys to adapt the training offer for the unemployed needs and offer them job counseling.
As the demand for jobs varies according to sectors evolution, the incorporation of technology and the geographical area evolve, it is necessary to be constantly monitoring this evolution in order to increase the supply of training or even to create new courses. It is intended, for example, that if tourism is growing in Alicante and companies in the sector invest in online business and demand professionals with specific knowledge, SERVEF can train the unemployed in those capacities.
The current prospecting process is expensive and slow compared to the pace of economic developments. In addition, the traditional method is based on interviews with workers, consultants and employers, obtaining exclusively qualitative information based on subjective opinions.
Using a technology used in Competitive Intelligence, SERVEF has made a pilot to obtain information on job portals, to classify these offers automatically, and to generate information to identify trends in the labor market.
The project has been developed by a mixed team of SERVEF and antara , company specialist in solutions for business intelligence. The antara team has built an automatic job classification system, with an expandable dictionary of 3,000 concepts and more than 700 "intelligence hypotheses". As a result, thousands of job offers have been classified based on a multitude of aspects, such as the region, the sector, the position within the company, if a specific degree is demanded or knowledge of which language, among many other factors.
The results of the pilot make it possible to obtain information to support decision-making within a week and by a single person, compared to the several months of work done by a few dozen people in the field in the current method .