Please enable JavaScript.
Coggle requires JavaScript to display documents.
Real-Time Big Data Models: Recommendation Systems (mohd. asil) - Coggle…
Real-Time Big Data Models: Recommendation Systems
(mohd. asil)
INTRODUCTION
Real-time personalization
Large-scale data handling
Tracks user behavior
Content-Based Recommendations
Uses item attributes (tags, genre)
Builds user profile from history
Similarity techniques (TF-IDF, Cosine)
Pros: Personalized, independent of other users
Cons: Limited discovery, needs item metadata
Collaborative Filtering
Based on user-item interaction
User-Based Filtering
Finds similar users
Item-Based Filtering
Recommends similar items
Pros: Learns patterns without item info
Recommends similar items
Cons: Cold start, sparse data issues
Real-Time Components
Data Ingestion: Apache Kafka
Stream Processing: Spark Streaming, Flink
Storage: MongoDB, Cassandra