Overview of Best Candidates

The AI Apps feature Best Candidates uses a machine learning model known as Intelligent Matching. The feature can be independently enabled. It helps you to quickly identify candidates that could be a good fit for a job requisition. Many jobs today receive hundreds of candidate applications, and it takes time for recruiters and hiring managers to find relevant candidates who’ve the right skills, experience, and education required for the job. The Best Candidates feature quickly compares and analyzes the details of candidate profiles with the details of an open job requisition. It then finds relevant candidates to be reviewed and saves time during the sourcing process.

The Best Candidates feature doesn’t replace the normal hiring process. It only provides recommendations. It’s up to you as the hiring manager or recruiter to decide which candidate is the best fit for the job through the hiring process.

Note: As with any artificial intelligence solution, the results shown are a machine’s best interpretation of the data. If data is missing from the candidate profile or job requisition, this will result in less relevant recommendations for that job and candidate.

Within Oracle Recruiting, the Best Candidate feature results are displayed within the Top Recommendations section of a job requisition. The recommendation is displayed along with a set of criteria, Profile, Education, Experience and Skills that help indicate why this candidate was recommended. This example shows that both candidates have a stronger match on the experience required for this job, but lower matches for skills, education, and profile.

How Best Candidate Works

The Best Candidates feature is powered by machine learning algorithms that evaluate candidate data and job requisition data to understand similarities across common words or phrases. The feature uses candidate and job requisition data and converts it into mathematical representations to calculate similarities. The feature is more powerful than a regular search because it uses Natural Language Processing (NLP) that can consider the context in which a word or phrase is located within the data. That means that the terms don’t have to be an exact match for the feature to find similarities in data.

Note: The Best Candidates feature doesn’t receive a copy of a candidate’s resume nor any PII data. Oracle Recruiting pushes a subset of fields about a candidate to AI Apps to use.

Considerations for Enabling Best Candidates

To use the Best Candidates feature, your organization must meet the following criteria:

• The Recruiting environment must not be on a EUR or government pod.

• The Recruiting environment must be live in production for at least 6 months, or migrated 6 months’ worth of data, to benefit from high-quality recommendations and prediction results.

• The Recruiting environment must use English for their operation language. If resumes, job applications, and job requisitions are in languages other than English, the candidate recommendations may be unpredictable. We recommend that you don't use this feature if you've data in languages other than English.
• The Best Candidate feature works best for job requisitions that require skilled labor (that is workers that have a specialized set of skills and education to perform the job).