What is a Résumé Screening System?
A résumé screening system automatically reads candidate résumés, extracts relevant information, and ranks applicants according to how closely they match a job description. The process replaces manual, time‑consuming review with reproducible, objective scoring.
Why Automate Résumé Screening?
- Speeds up hiring by processing hundreds of documents in seconds.
- Applies the same evaluation criteria to every candidate, reducing unconscious bias.
- Provides a transparent score that can be audited or adjusted.
- Scales easily as the volume of applications grows.
How to Build the System
- Set up the project: create a virtual environment, install dependencies (PyPDF2, python‑docx, pandas, Streamlit).
- Parse résumés: read PDF pages with PyPDF2 and DOCX paragraphs with python‑docx, normalise text to lowercase.
- Extract keywords: compare the résumé text against a skills taxonomy (JSON) using word‑boundary matching to avoid false positives.
- Design the scoring engine: assign weights to required skills, preferred skills, experience indicators, and additional keywords; compute a total score.
- Build the web interface: use Streamlit to accept a job description and multiple résumé uploads, trigger the backend, and display ranked results.
- Test and deploy: run the CLI or Streamlit locally, verify scores, then push to GitHub and deploy on Streamlit Cloud for public access.
Core Components Overview
- Résumé Parser – extracts plain text from PDF and DOCX files.
- Job Description Parser – identifies required and preferred skills.
- Keyword Extractor – matches résumé content to a predefined skills taxonomy.
- Scoring Engine – applies a weighted formula to generate a numeric rank.
- Streamlit Front‑end – provides a simple UI for input and result visualization.
Scoring Formula Explained
Total Score = (Required Skills × 0.50) + (Preferred Skills × 0.25) + (Experience × 0.15) + (Keywords × 0.10)
Each component is counted, multiplied by its weight, and summed to produce a score out of 100. Required skills carry the highest weight to ensure essential qualifications dominate the ranking.
Reducing Bias with Weighted Scoring
Because every résumé is evaluated against the same numeric criteria, factors such as formatting, writing style, or personal identifiers do not influence the outcome. The system therefore promotes a fairer, data‑driven short‑listing process.