A noticeable drop in your OneMatch score (e.g., 17 → 2) usually points to changes in your inputs or scoring configuration. Start by reviewing and comparing the previous vs. current OneMatch JSON to spot differences—especially in jobtitle, qualification, experience, and skills.
Common causes include:
- JD/resume updates: Modified titles, skills, or requirements reduce overlap.
- Parsing differences: New file formats, OCR/text quality, or language/locale settings changing field extraction.
- Configuration drift: Updated weights, must-have skills/thresholds, or taxonomy/synonym versions.
- Field mismatches: Qualification/education normalization or experience calculations (total vs. relevant) differing from earlier runs.
Quick checks
- Compare old vs. new JSON (same resume + JD): skills, titles, experience, education, locations, required/preferred.
- Confirm the same scoring profile (weights, thresholds, must-haves) and same taxonomy/synonym set.
- Verify parser version and locale; reparse originals if needed.
- Look for missing sections in the newer parse (e.g., empty skills) or problematic PDF-to-text conversions.
For a differential analysis, send the old & new OneMatch request and response and your current scoring/taxonomy settings to support@rchilli.com
Comments
0 comments
Please sign in to leave a comment.