Through the Hong Kong Qualifications Framework (HKQF), job-seekers' skills and job recruitment requirements are matched and scored. After multiple combined AI calculations, they can assist companies or organizations in selecting suitable employees. This research has won the HKICT Award "Best Commercial Value Award" and "Best Social Responsibility Award".
Project Background
The matching between companies recruiting suitable employees or talents choosing suitable corporate positions has always been a difficult human resource issue. If an enterprise chooses the wrong talents, not only the loss of capital will also bring about a halt to the development of the enterprise; and the selection of the wrong position for the talents which cannot use their abilities to the full. The mismatch between the two will also lose a lot of human resources and time.
The "Hong Kong Qualifications Framework-HKQF" established in 2008 attempts to subdivide all industries in Hong Kong into 7 competence and qualification levels, which can be used to standardize the competence of talents and the qualification requirements of industry positions. Therefore, our scientific research team believes that the "Hong Kong Qualifications Framework" can speed up and strengthen the matching accuracy between job applicant skills and job recruitment requirements.
Project Architecture
Competency Tree Data Model
We format the ability of talents and the qualification requirements of industry positions into a tree-like ability value (XML) consistent with the 7-level hierarchy of the "Hong Kong Qualifications Framework".
XML ability value required by industry position qualifications :
Talent ability XML ability value:
The Simple Role Assignment
Purpose: Find the most suitable talents for a position among multiple candidates
Algorithm: Tree Similarity Algorithm + Bonus Point Algorithm
The Group Role Assignment
Purpose: To find the most suitable talent for each position among multiple candidates and multiple positions
Algorithm: Modified Clustering and Means Algorithm, Negotiation Protocol and Collaborative Framework between Multi–agent servers and Team requester
System Testing and Experiments
Purpose: Reliability and Stableness of the assignment decision
If you are interested in our scientific research projects, please contact us