Core idea
Haerye keeps the long‑term direction clear while letting the app’s learning data decide what you should reinforce today.
- Weekly plan: the 40‑week phases decide what you learn next.
- Daily focus: mastery / heatmap / recommendations decide what you reinforce today.
- One rule: complete at least one recommendation card every day.
40‑week roadmap (kept as‑is)
Phase 1 (Weeks 1–8)
Goal: stabilize at ~50 points/day
Focus: Hangul foundation + beginner grammar + 100 core words
Phase 2 (Weeks 9–24)
Goal: stabilize at ~75 points/day
Focus: intermediate grammar + pronunciation focus + 300 words
Phase 3 (Weeks 25–40)
Goal: maintain 100 points/day
Focus: advanced grammar + AI conversation + full vocabulary completion
Daily 4‑step cycle (30–45 minutes)
1) Input (10 min)
Hangul or Vocabulary. Aim for 10–15 points.
- Hangul: Stroke order ON + Listen to pronunciation.
- Vocabulary: Romanization ON → gradually OFF.
- Prefer new items (avoid repeated tapping for points).
2) Judgment (10–15 min)
Flashcards or Grammar. This is where “weakness data” is created.
- Flashcards: 20 cards. “Know” means you can say it immediately.
- Grammar: 1 unit + read 2–3 examples out loud.
3) Output (5–10 min)
Pronunciation or Dictation. Turn input into usable output.
- Pronunciation: 5 attempts (target confidence ≥ 0.7).
- Dictation: 5–10 sentences.
4) Feedback (5 min)
Complete one recommendation card every day.
- Frequently confused words
- Weak grammar category
- Continue where you left off
Score strategy
Keep balance across modules. Avoid “all‑in” on a single module.
Phase 1
Hangul 15 + Vocab 10 + Flashcards 15 + Grammar/Pronunciation 10 ≈ 50
Phase 2
Hangul/Vocab 20 + Flashcards 25 + Grammar 15 + Pronunciation 15 ≈ 75
Phase 3
Hangul 15 + Vocab 15 + Flashcards 35 + Grammar/Pronunciation 35 = 100
Haerye‑only rules
- “Don’t know” in flashcards isn’t failure — it creates a map of weaknesses.
- Use the grammar heatmap: prioritize lighter categories first.
- For pronunciation, repeat the same sentence up to 3 times until you pass.
- Never skip recommendation cards — they are the personalization engine.