Upload 3 files

README.md

@@ -0,0 +1,48 @@
+---
+title: CPPS — Campus Placement Prediction System
+emoji: 🎓
+colorFrom: blue
+colorTo: purple
+sdk: gradio
+sdk_version: "6.13.0"
+app_file: app.py
+python_version: "3.10"
+pinned: false
+license: mit
+---
+
+# CPPS — Campus Placement Prediction System
+
+AI-powered resume-to-job-description matching system using shared production parsing, KeyBERT, and Sentence-BERT.
+
+## Features
+
+- **Resume Upload** — Supports PDF and DOCX
+- **Job Description Input** — Paste any JD text
+- **Placement Score** — Overall match percentage with level badge
+- **Component Breakdown** — 6-component MW-ESE scoring (Skills, Projects, Internships, Work Experience, Academics, Achievements)
+- **Shared Source of Truth** — `app_logic.py` + `cpps_core.py` power both the app and the notebooks
+- **Multi-Source Skill Mining** — Skills are collected from explicit skill sections, certifications, projects, internships, work experience, and full-text fallback parsing
+- **Bucketed JD Skills** - JD skills are split into `required`, `preferred`, and `optional` buckets
+- **Hybrid Weighted Skills Score** - `S_sk` combines bucket-weighted JD coverage, match confidence, and resume coverage with weights `0.45 / 0.35 / 0.20`
+- **Degree-Aware Academics** - Degree requirements in the JD activate academics scoring, with GPA/percentage normalization and degree fallback scoring
+- **Hybrid Internship Scoring** - Internships use count, semantic relevance, and parsed duration, including date ranges such as `Jan 2024 - Jun 2024`
+- **Relevant Certification Credit** - Certifications are scored by semantic relevance to the target JD instead of only flat counts
+- **Technical-Only Matching** - Soft skills are excluded so the matching pipeline focuses on hard technical skills
+- **Expanded Skill Recall** - Matching includes synonym expansion, compound JD phrase decomposition, lower adaptive thresholds, and ontology-cluster partial matches
+- **Match-Tier Debugging** - Debug output includes match types, JD bucket counts, source counts, rejected JD phrases, and diagnostic notes
+- **Improvement Simulation** — What-if scenarios ranked by score impact
+- **Actionable Recommendations** — Targeted suggestions to improve your profile
+
+## How It Works
+
+1. Upload your resume (PDF/DOCX)
+2. Paste the target job description
+3. Click **Analyze** to run the shared production pipeline and get the placement score plus detailed breakdown
+
+## Tech Stack
+
+- **NLP Models**: Sentence-BERT (`all-MiniLM-L6-v2`), KeyBERT, spaCy
+- **Skill Matching**: 5-pass (exact -> alias -> synonym -> adaptive semantic -> partial overlap/cluster)
+- **Scoring**: Multi-Weight Expert Scoring Engine (MW-ESE) with bucket-weighted hybrid `S_sk`
+- **UI**: Gradio

app.py

@@ -0,0 +1,461 @@
+"""
+CPPS Gradio App — Campus Placement Prediction System UI.
+
+Upload a resume (PDF/DOCX) and paste a Job Description to get:
+- Placement Score with match level
+- Component breakdown (6 MW-ESE components)
+- Improvement suggestions ranked by impact
+- Actionable recommendations
+"""
+
+import gradio as gr
+import traceback
+from app_logic import predict
+
+
+# ── Theme & CSS ──────────────────────────────────────────────────────────
+
+CUSTOM_CSS = """
+@import url('https://fonts.googleapis.com/css2?family=Inter:wght@400;500;600;700;800;900&display=swap');
+
+* { font-family: 'Inter', sans-serif !important; }
+
+.gradio-container {
+    max-width: 1200px !important;
+    margin: 0 auto !important;
+    background: linear-gradient(135deg, #0f0c29 0%, #1a1a3e 40%, #24243e 100%) !important;
+}
+
+/* ── Header ── */
+.app-header {
+    text-align: center;
+    padding: 2rem 1rem 1rem;
+    background: linear-gradient(135deg, rgba(99, 102, 241, 0.15) 0%, rgba(168, 85, 247, 0.12) 100%);
+    border-radius: 16px;
+    border: 1px solid rgba(99, 102, 241, 0.2);
+    margin-bottom: 1.5rem;
+}
+.app-header h1 {
+    font-size: 2rem;
+    font-weight: 800;
+    background: linear-gradient(135deg, #818cf8, #c084fc, #f472b6);
+    -webkit-background-clip: text;
+    -webkit-text-fill-color: transparent;
+    margin: 0 0 0.3rem;
+}
+.app-header p {
+    color: #94a3b8;
+    font-size: 0.95rem;
+    margin: 0;
+}
+
+/* ── Score Card ── */
+.score-card {
+    background: linear-gradient(135deg, rgba(99, 102, 241, 0.12) 0%, rgba(139, 92, 246, 0.08) 100%);
+    border: 1px solid rgba(99, 102, 241, 0.25);
+    border-radius: 16px;
+    padding: 2rem;
+    text-align: center;
+    margin-bottom: 1.5rem;
+    backdrop-filter: blur(12px);
+}
+.score-value {
+    font-size: 4rem;
+    font-weight: 900;
+    line-height: 1;
+    margin: 0.5rem 0;
+}
+.score-label {
+    font-size: 0.85rem;
+    text-transform: uppercase;
+    letter-spacing: 2px;
+    color: #94a3b8;
+    margin-bottom: 0.5rem;
+}
+.match-badge {
+    display: inline-block;
+    padding: 0.35rem 1.2rem;
+    border-radius: 20px;
+    font-weight: 700;
+    font-size: 0.8rem;
+    letter-spacing: 1px;
+    margin-top: 0.5rem;
+}
+.badge-excellent { background: rgba(52, 211, 153, 0.2); color: #34d399; border: 1px solid rgba(52, 211, 153, 0.3); }
+.badge-good { background: rgba(96, 165, 250, 0.2); color: #60a5fa; border: 1px solid rgba(96, 165, 250, 0.3); }
+.badge-moderate { background: rgba(251, 191, 36, 0.2); color: #fbbf24; border: 1px solid rgba(251, 191, 36, 0.3); }
+.badge-below { background: rgba(251, 146, 60, 0.2); color: #fb923c; border: 1px solid rgba(251, 146, 60, 0.3); }
+.badge-poor { background: rgba(248, 113, 113, 0.2); color: #f87171; border: 1px solid rgba(248, 113, 113, 0.3); }
+
+/* ── Section Card ── */
+.section-card {
+    background: rgba(30, 30, 60, 0.6);
+    border: 1px solid rgba(99, 102, 241, 0.15);
+    border-radius: 12px;
+    padding: 1.5rem;
+    margin-bottom: 1rem;
+    backdrop-filter: blur(8px);
+}
+.section-card h3 {
+    font-size: 1rem;
+    font-weight: 700;
+    color: #c4b5fd;
+    margin: 0 0 1rem;
+    display: flex;
+    align-items: center;
+    gap: 0.5rem;
+}
+
+/* ── Breakdown Table ── */
+.breakdown-table {
+    width: 100%;
+    border-collapse: separate;
+    border-spacing: 0 6px;
+}
+.breakdown-table th {
+    text-align: left;
+    font-size: 0.75rem;
+    text-transform: uppercase;
+    letter-spacing: 1px;
+    color: #64748b;
+    padding: 0.5rem 0.75rem;
+    font-weight: 600;
+}
+.breakdown-table td {
+    padding: 0.65rem 0.75rem;
+    font-size: 0.9rem;
+    color: #e2e8f0;
+    background: rgba(99, 102, 241, 0.05);
+    border-top: 1px solid rgba(99, 102, 241, 0.08);
+    border-bottom: 1px solid rgba(99, 102, 241, 0.08);
+}
+.breakdown-table tr td:first-child { border-left: 1px solid rgba(99, 102, 241, 0.08); border-radius: 8px 0 0 8px; }
+.breakdown-table tr td:last-child { border-right: 1px solid rgba(99, 102, 241, 0.08); border-radius: 0 8px 8px 0; }
+
+/* ── Progress bar ── */
+.progress-bar-bg {
+    background: rgba(99, 102, 241, 0.1);
+    border-radius: 6px;
+    height: 8px;
+    width: 100%;
+    overflow: hidden;
+}
+.progress-bar-fill {
+    height: 100%;
+    border-radius: 6px;
+    transition: width 0.6s ease;
+}
+
+/* ── Skill tags ── */
+.skill-tag {
+    display: inline-block;
+    padding: 0.25rem 0.65rem;
+    border-radius: 6px;
+    font-size: 0.8rem;
+    font-weight: 500;
+    margin: 0.2rem;
+}
+.skill-missing { background: rgba(248, 113, 113, 0.15); color: #fca5a5; border: 1px solid rgba(248, 113, 113, 0.2); }
+.skill-matched { background: rgba(52, 211, 153, 0.15); color: #6ee7b7; border: 1px solid rgba(52, 211, 153, 0.2); }
+
+/* ── Improvement row ── */
+.imp-row {
+    display: flex;
+    justify-content: space-between;
+    align-items: center;
+    padding: 0.6rem 0.75rem;
+    background: rgba(99, 102, 241, 0.05);
+    border: 1px solid rgba(99, 102, 241, 0.08);
+    border-radius: 8px;
+    margin-bottom: 6px;
+    font-size: 0.88rem;
+    color: #e2e8f0;
+}
+.imp-delta {
+    font-weight: 700;
+    color: #34d399;
+    white-space: nowrap;
+}
+
+/* ── Recommendation pill ── */
+.rec-item {
+    display: flex;
+    align-items: flex-start;
+    gap: 0.5rem;
+    padding: 0.5rem 0;
+    color: #cbd5e1;
+    font-size: 0.9rem;
+}
+.rec-bullet {
+    color: #818cf8;
+    font-weight: 700;
+    flex-shrink: 0;
+}
+
+/* ── Info cards row ── */
+.info-row {
+    display: grid;
+    grid-template-columns: 1fr 1fr;
+    gap: 1rem;
+    margin-bottom: 1rem;
+}
+.info-card {
+    background: rgba(30, 30, 60, 0.6);
+    border: 1px solid rgba(99, 102, 241, 0.15);
+    border-radius: 12px;
+    padding: 1.2rem;
+    text-align: center;
+}
+.info-card .label { font-size: 0.75rem; color: #94a3b8; text-transform: uppercase; letter-spacing: 1px; }
+.info-card .value { font-size: 1.5rem; font-weight: 700; color: #c4b5fd; margin-top: 0.3rem; }
+
+/* ── Error ── */
+.error-card {
+    background: rgba(239, 68, 68, 0.1);
+    border: 1px solid rgba(239, 68, 68, 0.3);
+    border-radius: 12px;
+    padding: 1.5rem;
+    color: #f1f5f9;
+}
+.error-card h3 { color: #ffffff; margin: 0 0 0.5rem; }
+.error-card pre { color: #e2e8f0; font-size: 0.8rem; white-space: pre-wrap; }
+
+/* ── Gradio overrides ── */
+footer { display: none !important; }
+"""
+
+THEME = gr.themes.Base(
+    primary_hue=gr.themes.colors.indigo,
+    secondary_hue=gr.themes.colors.purple,
+    neutral_hue=gr.themes.colors.slate,
+    font=gr.themes.GoogleFont("Inter"),
+)
+
+
+# ── HTML Rendering ───────────────────────────────────────────────────────
+
+def _score_color(score: float) -> str:
+    if score >= 85: return "#34d399"
+    if score >= 70: return "#60a5fa"
+    if score >= 50: return "#fbbf24"
+    if score >= 30: return "#fb923c"
+    return "#f87171"
+
+
+def _badge_class(level: str) -> str:
+    level_map = {
+        "EXCELLENT": "badge-excellent",
+        "GOOD": "badge-good",
+        "MODERATE": "badge-moderate",
+        "BELOW AVERAGE": "badge-below",
+        "POOR": "badge-poor",
+    }
+    return level_map.get(level, "badge-poor")
+
+
+def _bar_color(pct: float) -> str:
+    if pct >= 70: return "linear-gradient(90deg, #34d399, #6ee7b7)"
+    if pct >= 40: return "linear-gradient(90deg, #fbbf24, #fcd34d)"
+    return "linear-gradient(90deg, #f87171, #fca5a5)"
+
+
+def render_results(result: dict) -> str:
+    """Render pipeline result dict as styled HTML."""
+    score = result["placement_score"]
+    level = result["match_level"]
+    domain = result.get("domain_detected", "unknown").replace("_", " ").title()
+    role = result.get("role_title", "Not Specified")
+    color = _score_color(score)
+
+    # ── Score Card ──
+    html = f"""
+    <div class="score-card">
+        <div class="score-label">Placement Score</div>
+        <div class="score-value" style="color: {color}">{score}%</div>
+        <div class="match-badge {_badge_class(level)}">{level}</div>
+    </div>
+    """
+
+    # ── Info Row ──
+    html += f"""
+    <div class="info-row">
+        <div class="info-card">
+            <div class="label">Domain</div>
+            <div class="value">{domain}</div>
+        </div>
+        <div class="info-card">
+            <div class="label">Role</div>
+            <div class="value" style="font-size: 1.1rem">{role}</div>
+        </div>
+    </div>
+    """
+
+    # ── Component Breakdown ──
+    breakdown = result.get("component_breakdown", [])
+    excluded_names = {e["component"] for e in result.get("excluded_components", [])}
+    html += '<div class="section-card"><h3>📊 Component Breakdown</h3>'
+    html += '<table class="breakdown-table"><tr><th>Component</th><th>Weight</th><th>Score</th><th>Progress</th></tr>'
+    for item in breakdown:
+        name = item["component"]
+        total = item["total_score"]
+        achieved = item["score_achieved"]
+        pct = (achieved / total * 100) if total > 0 else 0
+        excluded = name in excluded_names
+        label = f'{name} <span style="color:#64748b;font-size:0.75rem">⊘ excluded</span>' if excluded else name
+        weight_pct = f"{total * 100:.1f}%"
+        score_pct = f"{achieved * 100:.1f}%"
+        bar = f"""<div class="progress-bar-bg"><div class="progress-bar-fill" style="width:{min(pct,100):.0f}%;background:{_bar_color(pct)}"></div></div>"""
+        opacity = "opacity: 0.4;" if excluded else ""
+        html += f'<tr style="{opacity}"><td>{label}</td><td>{weight_pct}</td><td>{score_pct}</td><td style="min-width:120px">{bar}</td></tr>'
+    html += '</table></div>'
+
+    # ── Skills Analysis ──
+    sa = result.get("skills_analysis", {})
+    matched_ct = sa.get("matched_count", 0)
+    total_ct = sa.get("total_jd_count", 0)
+    missing = sa.get("missing_skills", [])
+    html += '<div class="section-card"><h3>🎯 Skills Analysis</h3>'
+    html += f'<p style="color:#94a3b8;margin:0 0 0.8rem"><strong style="color:#e2e8f0">{matched_ct}/{total_ct}</strong> JD skills matched</p>'
+    if missing:
+        html += '<p style="color:#94a3b8;font-size:0.8rem;margin:0 0 0.5rem">Missing skills:</p><div>'
+        for skill in missing:
+            html += f'<span class="skill-tag skill-missing">{skill}</span>'
+        html += '</div>'
+    else:
+        html += '<p style="color:#34d399">✓ All required skills matched!</p>'
+    html += '</div>'
+
+    # ── Improvement Suggestions ──
+    suggestions = result.get("improvement_suggestions", [])
+    if suggestions:
+        html += '<div class="section-card"><h3>📈 Improvement Suggestions</h3>'
+        for s in suggestions[:8]:
+            html += f"""<div class="imp-row">
+                <span>#{s['rank']} {s['improvement']}</span>
+                <span><span style="color:#94a3b8;font-size:0.8rem">{s['score_change']}</span> &nbsp;<span class="imp-delta">{s['delta']}</span></span>
+            </div>"""
+
+        comb = result.get("combined_improvement", {})
+        if comb and comb.get("combined_delta", 0) > 0:
+            html += f"""<div style="margin-top:0.8rem;padding:0.8rem;background:rgba(52,211,153,0.08);border:1px solid rgba(52,211,153,0.2);border-radius:8px;text-align:center">
+                <span style="color:#94a3b8;font-size:0.85rem">Combined potential: </span>
+                <strong style="color:#34d399;font-size:1.1rem">{comb['current_score']}% → {comb['combined_new_score']}%</strong>
+                <span style="color:#34d399;font-weight:700"> (+{comb['combined_delta']}%)</span>
+            </div>"""
+        html += '</div>'
+
+    # ── Recommendations ──
+    recs = result.get("recommendations", [])
+    if recs:
+        html += '<div class="section-card"><h3>💡 Recommendations</h3>'
+        for rec in recs:
+            html += f'<div class="rec-item"><span class="rec-bullet">→</span><span>{rec}</span></div>'
+        html += '</div>'
+
+    return html
+
+
+def render_error(error_msg: str, tb: str = "") -> str:
+    """Render an error message as styled HTML."""
+    html = f'<div class="error-card"><h3>⚠ Error</h3><p>{error_msg}</p>'
+    if tb:
+        html += f'<details><summary style="cursor:pointer;color:#ffffff;margin-top:0.5rem">Technical Details</summary><pre>{tb}</pre></details>'
+    html += '</div>'
+    return html
+
+
+# ── Gradio Handler ───────────────────────────────────────────────────────
+
+def analyze(resume_file, jd_text):
+    try:
+        if resume_file is None:
+            return render_error("Please upload a resume file (PDF or DOCX).")
+        if not jd_text or not jd_text.strip():
+            return render_error("Please paste a job description.")
+
+        file_path = resume_file.name if hasattr(resume_file, "name") else str(resume_file)
+        result = predict(file_path, jd_text)
+        return render_results(result)
+
+    except Exception as e:
+        tb = traceback.format_exc()
+        return render_error(str(e), tb)
+
+
+# ── Build Gradio App ─────────────────────────────────────────────────────
+# Gradio 6: theme/css moved from gr.Blocks() constructor to app.launch()
+
+with gr.Blocks(
+    title="CPPS — Campus Placement Prediction System",
+) as app:
+
+    # Header
+    gr.HTML("""
+    <div class="app-header">
+        <h1>🎓 CPPS</h1>
+        <p>Campus Placement Prediction System — AI-Powered Resume Analyzer</p>
+    </div>
+    """, padding=False)
+
+    with gr.Row(equal_height=False):
+        # Left Column: Inputs
+        with gr.Column(scale=2):
+            resume_input = gr.File(
+                label="📄 Upload Resume",
+                file_types=[".pdf", ".docx"],
+                type="filepath",
+                elem_id="resume-upload",
+            )
+            jd_input = gr.Textbox(
+                label="📋 Job Description",
+                placeholder="Paste the full job description here...\n\nExample:\nSoftware Developer\n\nRequired Skills:\n- Python\n- SQL\n- Flask\n- REST APIs\n\n2+ years of experience\nB.Tech in Computer Science",
+                lines=12,
+                max_lines=25,
+                elem_id="jd-input",
+            )
+            analyze_btn = gr.Button(
+                "🔍  Analyze Match",
+                variant="primary",
+                size="lg",
+                elem_id="analyze-btn",
+            )
+
+        # Right Column: Results
+        with gr.Column(scale=3):
+            output_html = gr.HTML(
+                value="""
+                <div style="text-align:center;padding:4rem 2rem;color:#64748b">
+                    <p style="font-size:3rem;margin:0">🎓</p>
+                    <p style="font-size:1.1rem;margin:0.5rem 0 0.3rem;color:#94a3b8">Ready to Analyze</p>
+                    <p style="font-size:0.85rem">Upload a resume and paste a job description to get started</p>
+                </div>
+                """,
+                padding=False,
+                elem_id="results-output",
+            )
+
+    # Wire up
+    analyze_btn.click(
+        fn=analyze,
+        inputs=[resume_input, jd_input],
+        outputs=[output_html],
+    )
+
+    # Also trigger on Enter in JD box (submit event)
+    jd_input.submit(
+        fn=analyze,
+        inputs=[resume_input, jd_input],
+        outputs=[output_html],
+    )
+
+
+# ── Launch ───────────────────────────────────────────────────────────────
+# Gradio 6: theme and css are passed to launch(), not gr.Blocks()
+
+if __name__ == "__main__":
+    app.launch(
+        server_name="0.0.0.0",
+        server_port=7860,
+        theme=THEME,
+        css=CUSTOM_CSS,
+    )

requirements.txt

@@ -0,0 +1,15 @@
+gradio>=6.13.0
+
+transformers>=4.41.0
+torch>=2.0.0
+
+sentence-transformers>=2.7.0
+keybert>=0.8.0
+
+spacy>=3.7.0
+en-core-web-sm @ https://github.com/explosion/spacy-models/releases/download/en_core_web_sm-3.7.1/en_core_web_sm-3.7.1-py3-none-any.whl
+
+pymupdf
+python-docx
+numpy
+scikit-learn