Project

Back to overview

Cross-Media Indexing (XMI) for Multimedia Information Retrieval

English title Cross-Media Indexing (XMI) for Multimedia Information Retrieval
Applicant Crestani Fabio
Number 117994
Funding scheme Project funding (Div. I-III)
Research institution Istituto del Software (SI) Facoltà di scienze informatiche
Institution of higher education Università della Svizzera italiana - USI
Main discipline Information Technology
Start/End 01.05.2008 - 31.10.2011
Approved amount 149'122.00
Show all

Keywords (6)

Multimedia Information retrieval; Information retrieval; cross-media indexing; uncertainty modelling; information systems; Multimedia Information retrieval

Lay Summary (English)

Lead
Lay summary

The XMI project is investigating models of evidence combination, where evidence is heterogeneous and uncertain. These models can be applied to cross-media indexing as well as blog feed search, as the evidence combination problem can be formalised within the same modelling framework. During the XMI project, we experimented with machine learning techniques, smoothing methods, temporal analysis, and aggregation operators within the TREC blog track evaluation framework. Our experimental results are encouraging, improving on well established baselines, and suggest new directions of research that we are actively pursuing.

The work was carried out by the Principal Investigator together with Mostafa Keikha, the PhD student supported by the project. Mr. Keikha is working on the final part of his PhD thesis and is aiming to graduate in the next 6-7 months. 

Direct link to Lay Summary Last update: 21.02.2013

Responsible applicant and co-applicants

Employees

Publications

Publication
Relevance stability in blog retrieval
Keikha M, Gerani S, Crestani F (2011), Relevance stability in blog retrieval, in Proceedings of the ACM Symposium on Applied Computing, 1119-1123.
Time-based relevance models
Keikha M, Gerani S, Crestani F (2011), Time-based relevance models, in SIGIR'11 - Proceedings of the 34th International ACM SIGIR Conference on Research and Development in, 1087-1088.
Aggregating multiple opinion evidence in proximity-based opinion retrieval
Gerani S, Keikha M, Crestani F (2011), Aggregating multiple opinion evidence in proximity-based opinion retrieval, in SIGIR'11 - Proceedings of the 34th International ACM SIGIR Conference on Research and Development in, 1199-1200.
Blog distillation using random walks
Keikha M, Carman MJ, Crestani F (2009), Blog distillation using random walks, in Proceedings - 32nd Annual International ACM SIGIR Conference on Research and Development in Informat, 638-639.
Effectiveness of aggregation methods in blog distillation
Keikha M, Crestani F (2009), Effectiveness of aggregation methods in blog distillation, in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and , 5822 LNAI, 157-167.
Employing document dependency in blog search
Keikha M, Crestani F, Carman MJ, Employing document dependency in blog search, in Journal of the American Society for Information Science and Technology, 51(14).
Linguistic aggregation methods in blog retrieval
Keikha M, Crestani F, Linguistic aggregation methods in blog retrieval, in Information Processing and Management.

Scientific events

Active participation

Title Type of contribution Title of article or contribution Date Place Persons involved
Conference 18.04.2011 Dublin, Ireland
Conference 10.07.2010 Geneva, CH
Conference 16.10.2009 Roskilde, Denmark
Conference 17.09.2009 Milan, Italy
Conference 16.07.2009 Boston, USA
Spring School 14.05.2009 Mieres, Spain
Conference 10.04.2009 Toulouse, France
Conference 04.02.2009 Barcelona, Spain
Workshop 15.11.2008 Basel, CH


Self-organised

Title Date Place
Workshop 01.10.2008 Lugano

-