A method, apparatus, device, and storage medium for determining a course recommendation list.
By constructing a target-based basic data dictionary and sparse vectors, and combining multi-dimensional employee feature information, a course recommendation list is determined, which solves the problem of narrow recommendation scope in existing technologies and achieves higher quality and more accurate course recommendations.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- AGRICULTURAL BANK OF CHINA
- Filing Date
- 2022-10-20
- Publication Date
- 2026-05-26
AI Technical Summary
Existing course recommendation algorithms based on Bayesian probability statistics rely too heavily on existing user behavior data. As the number of executions increases, the range of recommended data decreases, making it easy to recommend only courses that users are interested in, rather than effectively recommending specific courses that match job requirements.
By acquiring sample data, a target basic data dictionary is constructed, resulting in a target sparse vector and an initial feature vector of the objects to be recommended. The target basic data dictionary and feature vectors are then used to determine the course recommendation list, taking into account the feature information of employees in various dimensions.
This improves the quality and accuracy of course recommendations, avoids scenarios where only courses of personal interest are recommended, and ensures that specific courses that match the job requirements are recommended.
Smart Images

Figure CN115563390B_ABST