# 3. Mastery Tracking

User Type: **Super Administrator**
Source: *Mi Digital Academy - Education CRM Features Document*

---

## 3. Mastery Tracking

### 3.1 Concept Mastery Visualization
**What it does:** Configures the concept mastery visualization: the mastery levels, the mastery display (per concept, per subject, per student), and the mastery trend view.

**Sub-features:**
- Mastery level definitions (e.g., New, Learning, Familiar, Mastered)
- Mastery display per concept, per subject, per student
- Mastery trend view (mastery over time)
- Mastery heatmap (concept × student or concept × subject)
- Mastery threshold per level (recall rate to reach each level)
- Available on web
- Event logging (action performed)
- Audit logging of mastery visualization configuration

**Super Admin User Journey:**
1. Super Admin opens SRS Management → Mastery Tracking → Visualization.
2. Defines the mastery levels (New, Learning, Familiar, Mastered) and the threshold per level.
3. Configures the mastery display per concept, per subject, and per student.
4. Enables the mastery trend view and the mastery heatmap.
5. Saves; the visualization reflects the configured levels and thresholds.
6. Each configuration change is event-logged and audit-logged.

**Rules & Edge Cases:**
- A concept's mastery level is derived from its recall rate against the level thresholds; there is no manual override.
- The mastery heatmap is computed from the same recall data as the per-concept view (no divergence).
- A concept with no recall data is shown as New (not an error).
- Configuration changes are audit-logged with the setting, the change, and the timestamp.

### 3.2 Confidence Rating Configuration
**What it does:** Configures the student confidence rating input used by the SRS: the rating options (Too Easy / Good / Hard), the rating prompt timing, and the rating-to-adjustment mapping.

**Sub-features:**
- Confidence rating options (Too Easy / Good / Hard)
- Rating prompt timing (after each recall, after a review batch)
- Rating-to-adjustment mapping (rating → interval adjustment)
- Rating required vs optional (per content type)
- Rating history per item
- Available on web
- Event logging (action performed)
- Audit logging of confidence rating configuration

**Super Admin User Journey:**
1. Super Admin opens SRS Management → Mastery Tracking → Confidence Rating.
2. Sets the confidence rating options (Too Easy / Good / Hard).
3. Configures the rating prompt timing (after each recall or after a review batch).
4. Maps each rating to its interval adjustment.
5. Sets whether the rating is required or optional per content type.
6. Saves; the rating input applies to SRS reviews.
7. Each configuration change is event-logged and audit-logged.

**Rules & Edge Cases:**
- A required rating blocks progression until the student provides a rating; an optional rating defaults to "Good" when skipped.
- Each rating produces exactly one interval adjustment; there is no compounding.
- The rating history per item is retained for analytics and is immutable.
- Configuration changes are audit-logged with the setting, the change, and the timestamp.

### 3.3 Retention Rate Prediction
**What it does:** Configures the retention rate prediction: the prediction model, the prediction horizon, and the at-risk concept identification.

**Sub-features:**
- Retention prediction model (based on recall history and Ebbinghaus parameters)
- Prediction horizon (e.g., 7-day, 30-day retention)
- At-risk concept identification (predicted retention below threshold)
- Prediction confidence (high/medium/low based on data volume)
- Prediction refresh frequency
- Available on web
- Event logging (action performed)
- Audit logging of retention prediction configuration

**Super Admin User Journey:**
1. Super Admin opens SRS Management → Mastery Tracking → Retention Prediction.
2. Configures the retention prediction model (recall history + Ebbinghaus parameters).
3. Sets the prediction horizon (7-day, 30-day retention).
4. Defines the at-risk concept threshold (predicted retention below which a concept is at-risk).
5. Sets the prediction confidence levels and the refresh frequency.
6. Saves; the prediction applies to SRS tracking.
7. Each configuration change is event-logged and audit-logged.

**Rules & Edge Cases:**
- A prediction is only shown when there is sufficient recall data; otherwise it is marked low-confidence or not shown.
- An at-risk concept is one whose predicted retention falls below the threshold; it is surfaced for reinforcement.
- The prediction is refreshed per the configured frequency; it is not computed on every view.
- Configuration changes are audit-logged with the setting, the change, and the timestamp.

### 3.4 Mastered Concept Graduation Rules
**What it does:** Defines the rules for graduating a mastered concept out of the active SRS queue: the graduation criteria, the graduation action, and the re-entry rule.

**Sub-features:**
- Graduation criteria (consecutive successful recalls, mastery level reached)
- Graduation action (remove from active queue, archive)
- Graduation review (periodic re-check of graduated concepts)
- Re-entry rule (a graduated concept re-enters on a failed recall)
- Graduation report (concepts graduated per period)
- Available on web
- Event logging (action performed)
- Audit logging of graduation rules

**Super Admin User Journey:**
1. Super Admin opens SRS Management → Mastery Tracking → Graduation Rules.
2. Sets the graduation criteria (consecutive successful recalls, mastery level reached).
3. Defines the graduation action (remove from active queue, archive).
4. Configures the periodic re-check of graduated concepts and the re-entry rule.
5. Enables the graduation report.
6. Saves; the rules apply to mastered concepts.
7. Each configuration change is event-logged and audit-logged.

**Rules & Edge Cases:**
- A concept graduates only when all graduation criteria are met; partial criteria do not graduate it.
- A graduated concept is removed from the active queue but retained in the archive for history.
- A graduated concept re-enters the active queue on a failed recall (re-entry rule).
- Configuration changes are audit-logged with the rule, the change, and the timestamp.