Enhanced Skill Recommendations Based on Nonskill Talent Profile Attributes

The Skill Similarity recommender has been enhanced to provide more relevant skill recommendations for employees, including organizations that are new to Skills or have limited skill data in talent profiles.

Previously, recommendations primarily used semantic similarity based on available employee-defined talent profile information from:

  • areas of expertise
  • areas of interest
  • career statement

With this enhancement, similarity matching can also use job and position context, including:

  • Work assignment name or business title
  • Job name (position name in case of positions)
  • Job profile name (position profile name in case of positions)

Recommendations continue to use any available profile, job, and position attributes when one or more attributes are missing.

The result of this change is that customers whose users have few or none of the employee defined values (area of expertise, area of interest, and career statement), will still be recommended relevant skills to add to their talent profile.

Employees receive recommendations in the Skills based on nonskill attributes on your talent profile recommendation category. Existing direct semantic-search behavior remains supported.

As in prior releases, this capability is delivered through an AI Studio Workflow Agent and includes a minimum similarity percentage (minimumSimilarityPercentage), which is set to 70. Customers can duplicate the agent and adjust the minimum similarity percentage threshold to increase or decrease the number of skill recommendations generated. Increasing the value makes the recommender more selective and may return fewer recommendations. Decreasing the value makes the recommender less selective and may return more recommendations.

Existing recommendation-generation and workflow-execution permissions continue to apply; no new access requirements are introduced.

Steps to enable and configure

You don't need to do anything to enable this feature.

Tips and considerations

To adjust the similarity threshold by duplicating the seeded Skill Recommendation Agent template and related Skills Library business object perform the following:

  1. Go to My Client Groups > AI Agent Studio > Marketplace.
  2. Search for and Open the Skill Recommendation Agent template.
  3. Copy the seeded agent template.
  4. Go to My Client Groups > AI Agent Studio > Resources > Business Objects.
  5. Copy the seeded Skills Library business object.
  6. In the copied Skills Library business object, update the minimumSimilarityPercentage value as needed and save it.
  7. Go back to AI Agent Studio > Workflows, search for the workflow you just created by copying the Skill Recommendation Agent template, replace the seeded Skills Library business object with the copied Skills Library business object in each BO function node that currently calls the Skills Library business object. While doing so, make sure that the parameter to the business object is retained as is.
  8. Save and publish the copied Skill Recommendation Agent.
  9. Set the value of the profile option ORA_HRT_SKILL_RECOMMENDATION_AGENT_CODE to the new agent you just published.