Spaces:
Running
Running
File size: 2,351 Bytes
d74863e b27958e d74863e | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 | # Recommendation Module 🎯
## Responsibility
This module handles **suggesting new educational videos** based on user interests.
## Functionality
1. Analyze user's saved notes.
2. Extract topics and categories.
3. Search **YouTube** for related videos.
4. Return a list of suggested educational videos.
## Files
### 1. `recommender.py`
- **Purpose:** Suggest educational videos from YouTube.
- **Main Class:** `RecommendationService`
- **Key Methods:**
- `get_recommendations_for_user(user_id)` - Get general recommendations for user.
- `get_youtube_recommendations(query)` - Search YouTube based on keywords.
- `get_similar_notes(note_id)` - Get similar notes from the same category.
## How It Works
1. **Fetch Notes:** Read the user's last 5 saved notes.
2. **Extract Topics:** Use categories or titles.
3. **Enhance Search:** Add keywords like "educational", "tutorial", "lecture".
4. **Filter Results:** Select only embeddable videos.
## Proposed Enhancements
- [ ] Add caching for recommendations to reduce YouTube API calls.
- [ ] Use Machine Learning to improve recommendation accuracy.
- [ ] Add filter for video duration (avoid very long videos).
- [ ] Prioritize well-known educational channels.
## Testing
```python
from src.ai_modules.recommendation.recommender import RecommendationService
from src.db.database import get_session
recommender = RecommendationService()
# Get recommendations for user
async def test_recommendations():
async with get_session() as session:
recs = await recommender.get_recommendations_for_user(session, user_id=1)
for video in recs:
print(f"{video['title']} - {video['url']}")
```
## Libraries Used
- `google-api-python-client` - Search YouTube.
- `groq` - Extract topics via LLM.
## Important Notes
- YouTube API has a **daily quota limit**, use caching to reduce requests.
- Enhanced search adds "educational lecture tutorial" keywords for better educational results.
- System prioritizes videos from the same category.
## API Quota Management
It's recommended to cache results temporarily to avoid exhausting YouTube API quota:
```python
# Simple cache example
cache = {}
def get_cached_recommendations(query):
if query in cache:
return cache[query]
results = youtube_search(query)
cache[query] = results
return results
```
|