Digital Library

cab1

 
Title:      ASSESSING JOB MARKET DYNAMICS USING ELK STACK
Author(s):      Gabriel Silva, Mário Rodrigues, Marlene Amorim, Angélica Souza, Marta Dias and Armando Pinho
ISBN:      978-989-8704-23-8
Editors:      Pedro Isaías
Year:      2020
Edition:      Single
Keywords:      Natural Language Processing, Ontology, Data Overview, ESCO, ELK Stack, Job Market
Type:      Full
First Page:      91
Last Page:      98
Language:      English
Cover:      cover          
Full Contents:      click to dowload Download
Paper Abstract:      The adoption of digital technologies promises to accelerate the transformation and the agility of processes, work activities and revenue models. Yet, the promised gains come together with dramatic needs for qualified professionals who can effectively leverage the technology potential. Job contexts are being reshaped as new models for the interaction and integration of humans and technologies take shape. To increase the readiness of the job market in this fast-changing context it is important that all stakeholders - companies, professionals, policy makers - are aware of the job market dynamics and needs. This can be observed from the collection of job announcements, but its high volume requires effective tools for analyzing and simplifying it in order to draw timely and correct conclusions. ELK stack was used for dealing with the high volume of job announcements. ELK is a stable platform that can manage large quantities of data and the Kibana layer enables to rapidly explore data and create visualization dashboards. As job announcements have distinct formulations for similar roles, depending on the hiring company, this raises the necessity of establishing a common ground for comparing the job descriptions. In this work were mapped job descriptions to ESCO occupations. ESCO is an ontology published by the European Union and its occupations are job positions. Results show that the ELK stack is a suitable tool for providing a visual interpretation on the job market dynamics. Moreover, the first experiments using natural language processing techniques and machine learning algorithms revealed an accuracy over 0.9 in mapping job descriptions to ESCO occupations. This result is very promising and shows that ESCO a good candidate as common ground to enable comparison of job market dynamics for distinct environments.
   

Social Media Links

Search

Login