Inverse square rank fusion for multimodal search

André Mourão, Flávio Martins, João Miguel da Costa Magalhães

Research output: Chapter in Book/Report/Conference proceedingConference contribution

5 Citations (Scopus)

Abstract

Rank fusion is the task of combining multiple ranked document lists (ranks) into a single ranked list. It is a late fusion approach designed to improve the rankings produced by individual systems. Rank fusion techniques have been applied throughout multiple domains: e.g. combining results from multiple retrieval functions, or multimodal search where several feature spaces are common. In this paper, we present the Inverse Square Rank fusion method family, a set of novel fully unsupervised rank fusion methods based on quadratic decay and on logarithmic document frequency normalization. Our experiments created with standard Information Retrieval datasets (image and text fusion) and image datasets (image features fusion), show that ISR outperforms existing rank fusion algorithms. Thus, the proposed technique has comparable or better performance than existing state-of-the-art approaches, while maintaining a low computational complexity and avoiding the need for document scores or training data.
Original languageEnglish
Title of host publication2014 12th International Workshop on Content-Based Multimedia Indexing (CBMI)
PublisherIEEE Computer Society
ISBN (Electronic)978-1-4799-3990-9
DOIs
Publication statusPublished - 2014
Event12th International Workshop on Content-Based Multimedia Indexing (CBMI 2014) - Klagenurt, Austria
Duration: 18 Jun 201420 Jun 2014
Conference number: 12th

Publication series

NameInternational Workshop on Content-Based Multimedia Indexing
PublisherIEEE Computer Society
ISSN (Electronic)1949-3991

Conference

Conference12th International Workshop on Content-Based Multimedia Indexing (CBMI 2014)
Abbreviated titleCBMI 2014
CountryAustria
CityKlagenurt
Period18/06/1420/06/14

Keywords

  • Better performance
  • Document frequency
  • Fusion algorithms
  • Fusion techniques
  • Individual systems
  • Low computational complexity
  • Multimodal search
  • State-of-the-art approach

Fingerprint Dive into the research topics of 'Inverse square rank fusion for multimodal search'. Together they form a unique fingerprint.

  • Cite this

    Mourão, A., Martins, F., & Magalhães, J. M. D. C. (2014). Inverse square rank fusion for multimodal search. In 2014 12th International Workshop on Content-Based Multimedia Indexing (CBMI) [6849825] (International Workshop on Content-Based Multimedia Indexing). IEEE Computer Society. https://doi.org/10.1109/CBMI.2014.6849825