Merging and Deduplication of Education/Work Experience Items🔗
Starting with release 6.17.4 (Beta), we’ve updated how candidate education and work history are processed during indexation. We’ve deprecated attempts to merge ATS (Applicant Tracking System) data with the information extracted from candidate resumes. This change was made because de-duplication and merging of education and work history items is not technically feasible. Previous attempts were unreliable and resulted in poor match quality.
To facilitate the new approach, five new fields have been introduced in the candidate summary to capture the relevant information, and enhance match accuracy. These new fields are:
- Most recent job title
- Most recent employer
- Start date of most recent employment
- Total years of experience
- Highest education level obtained
To optimize match results, it’s highly recommended to map these fields in the Candidate Mapping page as described here.
Fields will be computed as follows:
- Most recent job title: Pulled from the newly added Most Recent Job Title field (if mapped). If not available, it will be calculated from the ATS work history (if available and mapped), and if neither is available, it will be derived from the work history extracted from the candidate’s resume.
- Most recent employer: Pulled from the Most Recent Employer field (if mapped). If not available, it will be calculated from the ATS work history (if available and mapped), and if neither is available, it will be derived from the resume's work history.
- Total years of experience: Pulled from the Years of Experience field (if mapped). If not available, it will be calculated from the ATS work history (if available and mapped), otherwise derived from the work history extracted from the resume.
- Highest education level: Pulled from the Highest Education Level field (if mapped). If not available, it will be calculated from the ATS education history (if available and mapped), and if neither is available, it will be derived from the education details extracted from the resume.
Info
With release 6.17.4, these changes are in beta, so 2 custom settings must be set to activate the features. Under "Textkernel Indexing Service Settings", set "Use TMF Version 2" checkbox to "True", and change "Custom Indexation User Name" from the default value "import" to be "importv2". After this, all the newly-indexed or updated candidates in the Textkernel Search index will automatically benefit from the improved data handling.
In the next release, the default behaviour will change, so that these custom settings are removed and all customer environments will adopt the new behaviour.
If you prefer Textkernel to prioritize resume data over ATS data, simply remove the mapping for these fields on the Candidate Mapping page. This will ensure that Textkernel defaults to using the information extracted directly from the candidate’s resume.