Enabling multi-level relevance feedback on PubMed by integrating rank learning into DBMS
2010

RefMed: A New System for Finding Relevant Articles on PubMed

publication Evidence: high

Author Information

Author(s): Yu Hwanjo, Kim Taehoon, Oh Jinoh, Ko Ilhwan, Kim Sungchul, Han Wook-Shin

Primary Institution: POSTECH, Pohang, South Korea

Hypothesis

Can a multi-level relevance feedback system improve the accuracy of article retrieval on PubMed?

Conclusion

RefMed is the first multi-level relevance feedback system for PubMed, achieving high accuracy with less user feedback.

Supporting Evidence

  • RefMed integrates RankSVM into a DBMS to improve processing time.
  • The system allows users to provide multi-level feedback, enhancing relevance learning.
  • RefMed achieves high accuracy with less user feedback compared to traditional methods.

Takeaway

RefMed helps people find the right articles on PubMed by letting them give feedback on what they like, making it smarter over time.

Methodology

RefMed uses RankSVM for learning relevance from user feedback and integrates it into a relational database management system for real-time processing.

Limitations

The system's performance may depend on the amount and quality of user feedback provided.

Digital Object Identifier (DOI)

10.1186/1471-2105-11-S2-S6

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication