c0c0n3 / kitt4sme.live

On a mission to bring AI to the shop floor: https://kitt4sme.eu/
MIT License
1 stars 28 forks source link

Worker Profiler (T5.2) #92

Closed kostasgrevenitis closed 2 years ago

kostasgrevenitis commented 2 years ago

Worker Profiler

Task 5.2 – Human competence and capabilities profiling [Leader: GATE, Duration: M11 – M22]

This task develops the KITT4SME Worker Profiler, a tool that allows performing a skills gap assessment for the single worker to be used as an input of the Competence Matchmaking Tool (T5.5) that finds the optimal training plan to fill those gaps. Three steps are foreseen to develop the Worker Profiler.

First, inputs from T5.1 are used to identify both the new archetypes requested by the on-going digital transformation and the updated skills to be integrated in the existing archetypes to update them according to the new technological requirements. Second, a map of the skills mismatching is created with a questionnaire.

The questionnaire will be generated automatically using an AI based approach applied on the information extracted in the previous point. The tool (offered by GATE) will be used in order to make the questionnaire design phase more efficient and replicable. Finally, reference standards like ESCO and O*NET are used to quantify gaps and to create a proper assessment procedure to be integrated into the tool.

High-level functionality

The worker profiler is a tool able to assess the skill gap of a worker in relation to:

Profilers

Role in the architecture

Worker profiler is a platform application service. It is a RESTful backend service with an UI, that users use it to fill questionnaires based on their skills. It's output is the input for GATE's "Catalogue of Qualified Training Experiences" and "Competence Matchmaking Tool" tools.

Requirements

Improvements

It is its first implementation

Value proposition

If the worker is interested to measure the skill gap between what they currently do and how they want to improve, Worker Profiler gives the opportunity to select any archetype from a long list of them. Then, the worker conducts the assessment. The system takes the answers provided by the user and calculates a “similarity score” for every archetype. Finally, the three archetypes with the highest score are selected to be presented as suggestions to the user.

c0c0n3 commented 2 years ago

Content migrated to:

Please open a PR over there for further contribs.