How jatharthan/resume-to-job-matcher-spacy- works
This is a notebook-derived spaCy prototype for reciprocal resume/job recommendations. The user-facing workflow extracts skills from a resume or job description, compares them with the opposite-side collection, and ranks matches; the script also converts annotated data, trains and evaluates a named-entity model, and loads saved model artifacts. The main implementation is concentrated in final_research.py. The README describes the recommendation goal, but the excerpts do not show a deployed interface or a separate production service.
Main entry points: final_research.py Research script · model2 Saved spaCy model · traindata_3withmyannotation.json Annotated examples · 3class_test_data.json Test examples
Components 17
NER training
- Annotation conversion
final_research.py - Span cleanup
final_research.py - NER training
final_research.py - Model evaluation
final_research.py - Saved spaCy model
five_class_entitydata/model2
Matching workflow
- Research script
final_research.py - Skill extraction
final_research.py - Job collection
final_research.py - Resume collection
final_research.py - Match ranking
final_research.py - Ranked matches
final_research.py
Input corpora
- Annotated examples
five_class_entitydata/traindata_3withmyannotation.json - Test examples
five_class_entitydata/3class_test_data.json - Job descriptions
five_class_entitydata/jobs - Resume texts
five_class_entitydata/cv
Other components
- Job seeker
- Employer
Connections 23
- Annotated examples → Annotation conversion: reads
- Annotation conversion → Span cleanup: formats
- Span cleanup → NER training: supplies spans
- NER training → Saved spaCy model: saves model
- Test examples → Model evaluation: supplies examples
- Saved spaCy model → Model evaluation: loads model
- Research script → Annotation conversion: calls
- Research script → NER training: calls
- Research script → Model evaluation: calls
- Research script → Skill extraction: calls
- Job descriptions → Job collection: reads
- Resume texts → Resume collection: reads
- Research script → Resume collection: builds
- Research script → Job collection: builds
- Resume collection → Skill extraction: extracts skills
- Research script → Match ranking: calls
- Match ranking → Skill extraction: extracts skills
- Match ranking → Job collection: compares skills
- Match ranking → Resume collection: compares skills
- Match ranking → Ranked matches: returns scores
- Job seeker → Research script: provides resume
- Employer → Research script: provides job text
- Research script → Ranked matches: prints matches