1. Contents-based RS

Contensts-based Recommendation Systems
과거에 사용자가 선호했던 것과 유사한 item을 추천하는 모델. (The system learns to recommend items that are similar to the ones that the user liked in the past.)

1) Algorithm

1. Extract features
   Unstructured items를 structured item representation으로 변환하는 전처리 과정
2. Compute similarity
   Item representation 간의 유사도를 계산
   • Similarity: Euclidean, cosine, inner product, Pearson, Jaccard, etc

---

Reference

1. [토크ON세미나] 추천시스템 분석 입문하기
2. Recommendation Systems Handbook(Francesco Ricci et al.)