Similar Job Recommendations allows the customer to return a set of jobs that are similar to a job of interest to a job seeker.
A caller passes a
Job entity X, and an optional
resultSize number N, to the
API. The Cloud Job Discovery service returns up to N similar recommended jobs.
The results are based on:
- User behavior: API returns jobs viewed or applied for by other users who also viewed or applied for the passed-in job X
- Job metadata: API returns jobs with a similar classification to the passed-in job X as determined by the service's job classification machine learning model
The list of returned jobs is sorted by relevance (job N has higher relevance than job N+1).
De-Duplication: The service removes jobs from the results that are possible
duplicates of the passed-in job X. De-duplication logic is based on the passed-in
The following steps are necessary or recommended when using this feature:
- For each Search and Histogram request, the
userIdin the requestHeader/requestMetadata must be populated
- For each Search request, User Behavior Event reporting (via Pub/Sub)
must be implemented, and, if applicable, the implementation must send at least the
for each search request:
- and one of
- For best results, greater than 20,000 unique user queries per month and a large number of jobs per company are recommended
Recommended jobs are only returned for the company associated with the passed-in
job (jobs associated with the passed-in Job's