Skip to content
An official website of the OECD. Find out more
Created by the Public Governance Directorate

This website was created by the OECD Observatory of Public Sector Innovation (OPSI), part of the OECD Public Governance Directorate (GOV).

How to validate authenticity

Validation that this is an official OECD website can be found on the Innovative Government page of the corporate OECD website.

“The Work”, AI Job Recommendation Service Using the National Job Information Platform

Main screen of The Work, providing personalized job information

It is difficult for job seekers to identify the employment information that applies to them. "The Work" is a service provided through 'Work-Net', Korea's public jobs portal network. The service uses AI technology to analyze job seeker data and to provide employment infortmation that is tailored to each user. The service offers recommended information to job seekers in six job-related categories. By using the service, the time it takes to search for the information is reduced drastically.

Innovation Summary

Innovation Overview

It is extremely easy to find information about countless job opportunities around the world, through various online job portals. However, in this flood of information, it is not easy for job seekers to find information on suitable jobs. Job seekers need to set numerous search conditions to find the right job openings and search through multiple sites to find work-related training, certificates, and policies, etc. This means that the job-seeking process is very time-consuming.

To address this problem, the Korea Employment Information Service(KEIS) established “The Work,” a service that recommends jobs suitable for each job seeker without the need to conduct individual searches. Registered users of The Work simply need to log into the public jobs portal 'Work-Net' to receive customized 6 job-related information recommendations. The Work is the first public job portal serviced by an OECD member state to provide such comprehensive job information every day. The job opening recommendations are made using Artificial Intelligence(AI), which analyzes the job seeker’s CV registered on Work-Net, his/her past training experience, any funding or grants already received, and the job seeker’s own areas of interest.

The Work aims to do the following:
1) Reduce the amount of time it takes for job seekers to find information on the jobs they are actually looking for
2) Realize social value by supporting the employment of various classes including middle-aged persons, and women who've experienced breaks in their careers
3) Innovate data-based public services by boosting the use of public data

Various job seekers, including university graduates, women with career breaks, and middle-aged people, have received job recommendations and successfully found employment through this innovative service, which eliminates the need to engage in boring and difficult job searches. Since The Work is conveniently featured on the home screen of Work-Net, users do not have to navigate complex menus to gain access to the service. The service drastically reduces the time it takes for job seekers to perform information searches. Whereas job seekers previously had to search different sites for information on jobs, training, certificates, and relevant policies, The Work collects this information and recommends it to users, organizing all the information into six major job-related areas.

The Work pulls information from the 'National Job Information Platform', which boasts massive amounts of accumulated job information. The National Job Information Platform is a data innovation platform implemented by the Korean government that aims to improve the utilization of public data through the consolidation of job-related information. Some of the data offered on the platform includes information on job openings, training, and employment insurance and certificates, all of which was previously scattered among various public administration systems. The platform aims to help create jobs and match jobs with appropriate candidates. It is also significant in that it is the product of inter-agency collaboration and draws on data from both the KEIS and its partner agencies. The National Job Information Platform uses Machine Learning technology to collect and analyze the competencies of job seekers and the competencies sought after by enterprises seeking to hire. A new 'Job Competency-based Matching System' that goes beyond passive job recommendations and actively matches job seekers with companies will begin operation in 2020.

Innovation Description

What Makes Your Project Innovative?

The Work is innovative for the following reasons:
1) The service utilizes AI and Big Data technology. This means that large amounts of data can be utilized to produce new, meaningful information. The system’s use of AI and Big Data allows it to provide users with information in 6 key job-related areas, and to recommend jobs based on the job seeker’s location, preferences, or other criteria.

2) The service provides highly reliable information based on the National Job Information Platform. The National Job Information Platform is a consolidated database for public administration data, including information on job openings, training, and certificates, etc. The quality of this information far surpasses that of the information available to private businesses, whose data is limited to their own clients.

3) The service is the first of its kind to be offered through a public portal of an OECD member state, providing automatic recommendations for a range of information.

What is the current status of your innovation?

The KEIS is engaged in the following three efforts to further improve The Work.
1) Through Work-Net log analysis, the degree of utilization and accuracy of the recommended information is being analyzed. Some of the items currently being analyzed are the degree to which users view the recommended information, and the degree to which the information recommended leads to the actual submission of job applications.

2) Algorithms that recommend job seekers to businesses looking to hire are currently being developed. This service is intended to recommend suitable candidates to businesses seeking to hire new employees.

3) Employment Insurance data is being analyzed to monitor the degree to which job seekers are successfully finding employment at the companies recommended to them, and the degree to which they are maintaining their employment after being hired. Algorithms are also being developed that can recommend jobs with high potential employee retention rates.

Innovation Development

Collaborations & Partnerships

The Work is the outcome of collaboration between numerous partners.
1) Interviews were conducted with citizens including job seekers, and business owners, and the information extracted from the interviews was analyzed.

2) Reliable public data and outside data was received from partner agencies and institutions, including government ministries and agencies, and private organizations.

3) Major Korean IT companies were consulted to establish the service’s AI technology and Big Data analysis.

Users, Stakeholders & Beneficiaries

The Work aims to help various job seekers, including university graduates, women with career breaks, and middle-aged people, gain access to more information on the jobs suited to their competencies and ultimately gain employment. The Work also connects companies seeking to hire new employees. For job seekers, The Work provides information in 6 key job-related areas, meaning that job seekers no longer have to separately search for information by visiting individual websites.

Innovation Reflections

Results, Outcomes & Impacts

Since its launch in December 2018 until up to November 2019, The Work has helped some 7,600 individuals find jobs well-suited to their unique competencies. This number was obtained by tracking the number of job seekers who applied for employment with companies recommended by The Work, were hired, and successfully gained employment insurance coverage. An analysis showed that The Work users who successfully found employment through the service variously ranged in age from their 20s to their 60s. Other analysis showed that while job seekers had previously spent an average of 10 minutes searching for job-related information on other, separate sites, The Work provided the same information within 5 seconds of log-in. In short, The Work reduces the amount of time people spend searching to approximately 1/120.

Challenges and Failures

The success of The Work is based largely on the amount and quality of the data available on the National Job Information Platform for AI and Big Data analysis. To ensure data volume and quality, the platform was initially designed to not only pull data from the KEIS's services, but also from other agencies under the Ministry of Employment and Labor. However, this data was still not enough for the service to provide job seekers with relevant information. Attempts were then made to obtain data from other agencies and institutions, such as government ministries, public agencies, and others, but there were no legal grounds in place for the furnishing of such information. To address this problem, the KEIS and the Ministry of Employment and Labor proposed an amendment to the 'Framework Act on Employment Policy' to allow for the furnishing of data by outside bodies. The amendment was signed into law in April 2019, allowing for outside data to be included in The Work system.

Conditions for Success

The success of The Work is based on the following:
1) The Work utilizes vast amounts of public data accumulated by the National Job Information Platform. The public data is more accurate than the data, statistics, and survey data collected by private corporations, resulting in data analysis results that are much more reliable.

2) The Work was made possible by KEIS CEO's leadership. An amendments to related laws were implemented by CEO's dedication to allow for the use of external data. The CEO made an effort for organization restructuring as well to ensure the implementation of the service included the operation of a specialized task force and dedicated department for service implementation within the KEIS.

3) The Work is the outcome of collaboration between numerous partners. Opinions were gathered from the Korean people, who are the intended end-users of The Work. Major Korean IT companies and research institutions were consulted on AI technology and Big Data analysis.

Replication

The Work can potentially be replicated in and outside of Korea. The public data of the National Job Information Platform will soon be made available to jobs portal network of Korean private sector, allowing the private sector to provide high-quality job services similar to those offered by The Work. The KEIS is currently involved in implementing computerized employment service networks in Asian countries such as Mongolia, Vietnam, and Cambodia through international cooperation programs. The Work has been a topic of keen interest for many of these countries. In the future, the Work is expected to continue to be benchmarked as an example of public employment services by countries around the world.

Lessons Learned

Data quality is the most important aspect of AI and Big Data analysis. Data quality can be improved through procedures such as data cleansing, which is an extremely important area in the field of data science. If the data used for analysis is poor in quality and has not been properly cleansed, it doesn’t matter how advanced the data analysis method is-the analysis results can still be biased with distorted conclusions. The cost, time, and manpower it takes to identify and correct the results of an improper data analysis are far greater than those needed to correctly cleanse data prior to analysis. Therefore, efforts are needed to secure reliable data and to improve the accuracy of the data input by the user. That is, it is necessary to recognize the importance of ensuring data quality, and to establish effective strategies for data quality control. In order to create value in public services through Big data, it is important to establish consistent objectives, strict processes to guarantee data accuracy, and strategic plans that address the attitudes of the personnel handling the data.

Anything Else?

The KEIS is seeking to improve The Work by adding a job competency-based matching algorithm. This improved service is scheduled to begin operation in June 2020. The service will use CVs from job seekers and job openings posted by hiring companies to create a Job Data Dictionary that matches candidates and companies by analyzing job suitability. This is expected to help reduce mismatches between job seeker competencies and the competencies sought by hiring companies, while at the same time improving the efficiency of the labor market and increasing employee retention rates.

Project Pitch

Year: 2018
Level of Government: National/Federal government

Status:

  • Developing Proposals - turning ideas into business cases that can be assessed and acted on
  • Implementation - making the innovation happen
  • Evaluation - understanding whether the innovative initiative has delivered what was needed

Innovation provided by:

Media:

Date Published:

5 August 2020

Join our community:

It only takes a few minutes to complete the form and share your project.