"The Work" is an AI-enabled job recommendation service developed by the Korea Employment Information Service (KEIS), operating under the Ministry of Employment and Labor (MOEL). Launched in December 2018, the service is delivered through Work-Net, Korea's national public jobs portal, and provides registered users with personalised daily recommendations across six job-related categories upon logging in.
The system uses Artificial Intelligence and Big Data technology to analyse each registered job seeker's CV, past training experience, funding and grants received, and stated areas of interest. Based on this analysis, it generates tailored recommendations covering job openings, vocational training, funding and grants, employment programmes, certificates, and employment insurance information. The service eliminates the need for job seekers to manually search across multiple websites and set numerous search conditions to find relevant opportunities.
The Work draws its data from the National Job Information Platform, a consolidated government data innovation platform that aggregates job-related public administration data previously scattered across various agencies. The platform includes information on job openings, training, employment insurance, and certificates, sourced from both KEIS and its partner agencies through inter-agency collaboration. The National Job Information Platform uses Machine Learning technology to collect and analyse the competencies of job seekers and the competencies sought by enterprises seeking to hire. A 'Job Data Dictionary' was constructed from 18 job-related datasets and approximately 2.7 million extracted keywords to support competency-based matching.
The quality and volume of data available for AI and Big Data analysis was a central challenge during development. The platform was initially designed to pull data from KEIS's own services and other agencies under the Ministry of Employment and Labor, but this proved insufficient. Attempts to obtain data from additional government ministries, public agencies, and other institutions were hampered by the absence of legal grounds for such data sharing. To address this, KEIS and MOEL proposed an amendment to the Framework Act on Employment Policy, which was signed into law in April 2019, enabling outside bodies to furnish data to the system.
The service targets diverse job seeker populations, including university graduates, women with career breaks, and middle-aged persons, aiming to reduce information search time and support employment across these groups. The Work is described as the first public job portal service offered by an OECD member state to provide comprehensive automatic daily recommendations across multiple job-related information categories.
The development of The Work involved collaboration with multiple stakeholders. Interviews were conducted with citizens including job seekers and business owners, and the resulting information was analysed to inform service design. Reliable public data and external data was received from partner agencies and institutions including government ministries and private organisations. Major Korean IT companies were consulted to establish the service's AI technology and Big Data analysis capabilities. Research institutions were also consulted on AI technology and Big Data analysis.
KEIS leadership played a significant role in enabling the service, including driving the legal amendments for external data use and restructuring the organisation to establish a specialised task force and dedicated department for service implementation within KEIS.
Regarding outcomes, from December 2018 to November 2019, approximately 7,600 individuals gained employment through companies recommended by The Work. This figure was verified by tracking job seekers who applied to recommended companies, were hired, and successfully gained employment insurance coverage. Users who found employment through the service ranged in age from their 20s to their 60s. Analysis showed that whereas job seekers previously spent an average of 10 minutes searching for job-related information across separate sites, The Work provided the same information within 5 seconds of logging in — reducing search time to approximately 1/120 of the previous duration.
At the time of the OECD OPSI case study, KEIS was engaged in ongoing improvement efforts: analysing Work-Net log data to assess utilisation and accuracy of recommendations; developing algorithms to recommend job seekers to businesses seeking to hire; analysing Employment Insurance data to monitor employment outcomes and retention rates at recommended companies; and developing algorithms to recommend jobs with high potential employee retention rates. A Job Competency-based Matching System that would actively match job seekers with companies was scheduled to begin operation in 2020.
Data quality was identified as the most critical factor for AI and Big Data analysis. KEIS emphasised the importance of data cleansing procedures and strategies for data quality control, noting that poor-quality data produces biased and distorted results regardless of the sophistication of the analysis method.