The recommender takes user input in the form of a job category and a job title and matches this to an ESCO role which is then matched to a suitable recruiter. The suitability of a recruiter is evaluated through its previous assignments as well as their field of specialty.
This test environment contains six recruiters: Team Interim, Digitalenta, Quadruped, Ants, Dreamwork, Closers.
Their recruiter profiles, as seen through the links above, have been matched to corresponding ESCO areas. For example, Ants have been mapped to ESCO codes starting with 25 (Information and communications technology professionals) and 215 (Electrotechnology engineers).
Each recruiter has also been assigned some dummy data as previous assignments according to the following table:
Recruiter | Role | ESCO code |
---|---|---|
Ants | ethical hacker | 2529.4 |
Dreamwork | ICT network architect | 2523.2 |
Quadruped | chief ICT security officer | 2529.1 |
Dreamwork | investment analyst | 2413.1.2 |
Team Interim | investment manager | 2412.6.3 |
For example, if I would like to recruit an IT security specialist, Ants and Quadruped would both be good recommendations as they have done similar assignments recently. Dreamwork could also be viable as they have a broad recruiter profile. Digitalenta and Team Interim should both be irrelevant as they do not have profiles related to IT.
The idea behind this matching algorithm is to later implement it as part of the request forms in order to be able to recommend recruiters without needing a manual mapping.
For more details view the GitHub repository here.