BACKEND ISSUE 03 — Frontier Detection Matrix

[LeDuyCoder]

🎯 Goal

Implement API trả về dữ liệu các chủ đề công nghệ mới nổi dựa trên Impact vs Velocity.

Reference UI: Frontier Detection Bubble Chart

---

📡 Endpoint

GET /analytics/frontier

---

🧾 RESPONSE CONTRACT

{
  "code": 200,
  "message": "Fetch frontier topics successfully",
  "data": [
    {
      "topic": "Gen AI",
      "impactFactor": 2.4,
      "citationVelocity": 4.5,
      "status": "EMERGING"
    },
    {
      "topic": "CRISPR V3",
      "impactFactor": 4.8,
      "citationVelocity": 8.9,
      "status": "FRONTIER"
    }
  ]
}

---

🧠 BUSINESS LOGIC

1. Data Model

Each topic contains:

• topic (string)
• impactFactor (float)
• citationVelocity (float)
• status (enum)

---

2. Status Classification Rules

• EMERGING:

  • impactFactor < 3
  • citationVelocity < 5

• FRONTIER:

  • impactFactor >= 3 AND citationVelocity >= 5

• DECLINING (future extension):

  • low velocity trend

---

3. Data Source (Phase 1)

• static mock dataset
• later replace with ML/scientometric model

---

⚠️ EDGE CASES

• missing values → exclude record
• negative values → clamp to 0
• invalid status → recompute server-side

---

📦 PERFORMANCE NOTES

• lightweight dataset
• suitable for caching
• expected < 100ms response time (mock phase)

---

🧪 ACCEPTANCE CRITERIA

• returns valid array of topics
• no null fields
• status always computed or validated
• FE can render bubble chart directly

---

🚀 FUTURE IMPROVEMENTS

• add time series for each topic
• integrate citation prediction model
• add filtering by domain