Zzong's Notes

Home

❯

papers

❯

recommender_system

❯

When Relevance is not Enough - Promoting Diversity and Freshness in Personalized Question Recommendation

When Relevance is not Enough - Promoting Diversity and Freshness in Personalized Question Recommendation

2026년 8월 29일1 min read

  • Paper link
    • http://www.pelleg.org/shared/hp/download/relevance-is-not-enough.pdf
  • 배경
    • question-answering sites 에 대한 추천 시스템
    • The feature that they really missed was diversity.
      • In other words, showing them just the main topics they had previously expressed interest in was simply too dull.

함께 보면 좋은 글

Joint User-Entity Representation Learning for Event Recommendation in Social Network

Paper link Abstract 사용자와 이벤트를 동일한 latent space 에 project 하기 위해 a joint representation model (결합 표현 모델) 을 제안 제안한 모델은 parallel CNN 기반이고, event text 의 semantic meaning 을 파악하며...

Practical Lessons for Job Recommendations in the Cold-Start Scenario

paper link .

Personalizing Software and Web Services by Integrating Unstructured Application Usage Traces

Tags paper review, implicit feedback, Recommendation System, WWW 17 paper link Abstract util2vec 을 제안: usage history, interaction records 를 활용 utilization-to-vector 제안한 방식의 세...

User Fatigue in Online News Recommendation

Abstract the impact of User Fatigue has been mostly ignored in the literature User fatigue 란, 특정 유저에게 동일한 아이템을 반복적으로 보여줬을 때, 해당 아이템에 대해 빠르게 흥미를 잃는 현상을 의미한다.

A Survey on Deep Learning Based POI Recommendations

A POI recommendation technique essentially exploits users’ historical check-ins and other multimodal information to recommend the next set of POIs suitable for a user.

Collaborative Topic Modeling for Recommending GitHub Repositories

References paper link: www.cs.cmu.edu/afs/cs.cmu.edu/Web/People/norii/pub/github-ctr.pdf .

Online learning to rank for information retrieval

References slide: staff.fnwi.uva.nl/m.derijke/wp-content/uploads/sigir2016-tutorial.pdf .

Modeling personalized item frequency information for next-basket recommendation

Abstract SBRS 에서 Next-basket 추천 (NBR) 을 위해 제안된 논문이다. RNN 모델이 sequential modeling 에 효과적이지만 추천 시나리오에서 item frequency 정보를 직접적으로 잡아낼 수 없다는 문제가 있다고 주장한다.

VBPR - Visual Bayesian Personalized Ranking from Implicit Feedback

Abstract 개인화 추천 및 ranking 방식을 진행할 때 아이템의 visual appearance 무시하는 경우가 많다.

Algorithmic Effects on the Diversity of Consumption on Spotify

Generalist Specialist Score (GS-score) 주어진 유저의 다양성을 수치적으로 측정하는 metric: GS-score 는 어떤 한 노래의 vector 와 사용자들의 vector 들 간 cosine similarity 의 평균을 측정한다.