BACKEND ISSUE 02 — Publication and Citation Historical Trends

[LeDuyCoder]

🎯 Goal

Implement API trả về dữ liệu biểu đồ xu hướng bài báo và trích dẫn theo thời gian.

Reference UI: Publication Trend + Citation Mirroring Chart

---

📡 Endpoint

GET /analytics/trends

---

🧾 RESPONSE CONTRACT

{
  "code": 200,
  "message": "Fetch publication trends successfully",
  "data": {
    "timeline": ["2021", "2022", "2023", "2024", "2025", "2026"],
    "series": [
      {
        "name": "Articles",
        "data": [12000, 15000, 18500, 22000, 28000, 32950]
      },
      {
        "name": "Citations",
        "data": [85000, 110000, 142000, 190000, 250000, 310000]
      }
    ]
  }
}

---

🧠 BUSINESS LOGIC

1. Timeline generation

• Years must be sorted ascending
• Must be consistent across all series

2. Series data

• Each series must match timeline length
• Missing values must be filled with 0

3. Data sources (phase 1)

• Mock dataset (in-memory)
• Can later be replaced by DB aggregation

---

⚠️ EDGE CASES

• Missing year → skip or fill 0 (must be consistent rule)
• Series mismatch → normalize length
• Empty dataset → return empty arrays safely

---

📦 PERFORMANCE NOTES

• Lightweight aggregation endpoint
• Suitable for caching (Redis later)
• Expected response time < 100ms (mock phase)

---

🧪 ACCEPTANCE CRITERIA

• timeline always sorted
• series length equals timeline length
• no null values in response
• FE can render chart directly without transform

---

🚀 FUTURE IMPROVEMENTS

• Add date range filter (7d / 30d / yearly)
• Add dynamic granularity (month/year)
• Replace mock with time-series DB aggregation