Youth employment

Meet our iTrack-Skills: Artificial Intelligence for tracing TVET students

How many TVET graduates found a stable job? Are they working in a job they are qualified for? How many graduates are unemployed or inactive? How relevant and effective was the training?

News | 22 January 2020

© iStock

Training schools and administrators of TVET (Technical and Vocational Education and Training) systems around the world are keen to know how graduates are doing on the labour market. One way of finding out are tracer studies and graduate surveys that analyse graduates’ employment status and work trajectories after completing a training programme. iTrack-Skills is an interactive tool for tracing graduates by making use of artificial intelligence and new digital tools. iTrack Skills was developed by the ILO and being piloted in Montenegro.

If systematically applied, tracer studies and graduate surveys can provide valuable intelligence to adjust and increase the relevance and quality of training programmes. They can become an important instrument within a system of different skills anticipation tools and can inform education planning and curriculum development within a larger methodological context. A large amount of literature is available on the topic, among them a practical guide jointly developed by ILO, ETF and Cedefop (which can be downloaded here). A major challenge, however, is to ensure a high participation of graduates in these surveys. Low response rates significantly limit the relevance and reliability of the findings.

This is where iTrack Skills comes in. This tool proposes to use artificial intelligence to appeal to the media and technology choices of the young generation. It is a text-messaging chatbot (a computer programme that simulates human conversations and allows interaction between a human and a machine) for gathering, processing and analysing information through interactive “chats”. Through pre-calculated automated responses, it simulates interactive human conversations based on pre-calculated responses, similar to the chatbots used in customer services. The underlying assumption is that the interaction with artificial intelligence is an exciting experience that young graduates will be interested to engage in.

iTrack-Skills will be piloted in three different countries: Montenegro, Tanzania and Bangladesh, primarily with graduates from TVET schools. In Montenegro, iTrack will use the questions of the regular tracer studies conducted by the Ministry of Education and Science. The Ministry faced problems in getting sufficient responses and the iTrack represents a welcome additional tool to diversify existing tracing tools and to increase the outreach to graduates.

iTrack-Skills will work as a web-based chat sending a URL via a messaging application that is popular among young people in Montenegro. It will lead the recipients to a sequence of questions that changes according to the individual user-response. Whenever a user replies to a question, iTrack will utilize artificial intelligence to read answers and choose new questions based on an algorithm. The information collected will be stored safely and confidentially in a database to which only the Ministry of Education and Science of Montenegro will have access.

An admin portal will manage the entire operation, including creating and sending surveys, as well as analysing and providing an overview over the responses. The analysis and final results will be displayed in a intuitive and graphical way. The Ministry will have a quick and easy way to receive and analyse graduate feedback in real time to understand educational trajectories and employment outcomes of TVET students. This information can be used for:

•    supporting the design, adjustment and updating of curricula
•    improving skills matching
•    educational planning and forecasting, and
•    strengthening career guidance. 

In addition to tracer studies, a tool like iTrack-Skills can be used to complement information needs in other policy areas and adjusted to the needs of the specific audience.