Beyond Shot Retrieval: Searching for Broadcast News Items Using Language Models of Concepts

Robin Aly; Aiden Doherty; Djoerd Hiemstra; Alan Smeaton

doi:10.1007/978-3-642-12275-0_23

Beyond Shot Retrieval: Searching for Broadcast News Items Using Language Models of Concepts

Robin Aly
, Aiden Doherty
, Djoerd Hiemstra
, Alan Smeaton

Databases (Former)

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

9 Citations (Scopus)

159 Downloads (Pure)

Abstract

Current video search systems commonly return video shots as results. We believe that users may better relate to longer, semantic video units and propose a retrieval framework for news story items, which consist of multiple shots. The framework is divided into two parts: (1) A concept based language model which ranks news items with known occurrences of semantic concepts by the probability that an important concept is produced from the concept distribution of the news item and (2) a probabilistic model of the uncertain presence, or risk, of these concepts. In this paper we use a method to evaluate the performance of story retrieval, based on the TRECVID shot-based retrieval groundtruth. Our experiments on the TRECVID 2005 collection show a significant performance improvement against four standard methods.

Original language	Undefined
Title of host publication	Proceedings of the 32nd European Conference on IR Research (ECIR 2010)
Editors	Cathal Gurrin, Yulan He, Gabriella Kazai, Udo Kruschwitz, Suzanne Little, Thomas Roelleke, Stefan Rüger, Keith van Rijsbergen
Place of Publication	London
Publisher	Springer
Pages	241-252
Number of pages	12
ISBN (Print)	978-3-642-12274-3
DOIs	https://doi.org/10.1007/978-3-642-12275-0_23
Publication status	Published - 2010
Event	32nd European Conference on Information Retrieval, ECIR 2010: (IR Resarch) - Milton Keynes, United Kingdom Duration: 28 Mar 2010 → 31 Mar 2010 Conference number: 32

Publication series

Name	Lecture Notes in Computer Science
Publisher	Springer Verlag
Volume	5993/2010
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	32nd European Conference on Information Retrieval, ECIR 2010
Abbreviated title	ECIR
Country/Territory	United Kingdom
City	Milton Keynes
Period	28/03/10 → 31/03/10

Keywords

IR-71235
EWI-17850
METIS-270807

Access to Document

10.1007/978-3-642-12275-0_23

Cite this

Aly, R., Doherty, A., Hiemstra, D., & Smeaton, A. (2010). Beyond Shot Retrieval: Searching for Broadcast News Items Using Language Models of Concepts. In C. Gurrin, Y. He, G. Kazai, U. Kruschwitz, S. Little, T. Roelleke, S. Rüger, & K. van Rijsbergen (Eds.), Proceedings of the 32nd European Conference on IR Research (ECIR 2010) (pp. 241-252). Article 10.1007/978-3-642-12275-0_23 (Lecture Notes in Computer Science; Vol. 5993/2010). Springer. https://doi.org/10.1007/978-3-642-12275-0_23

Aly, Robin ; Doherty, Aiden ; Hiemstra, Djoerd et al. / Beyond Shot Retrieval: Searching for Broadcast News Items Using Language Models of Concepts. Proceedings of the 32nd European Conference on IR Research (ECIR 2010). editor / Cathal Gurrin ; Yulan He ; Gabriella Kazai ; Udo Kruschwitz ; Suzanne Little ; Thomas Roelleke ; Stefan Rüger ; Keith van Rijsbergen. London : Springer, 2010. pp. 241-252 (Lecture Notes in Computer Science).

@inproceedings{01d4f6b227e245abb3404b517d7bc6b7,

title = "Beyond Shot Retrieval: Searching for Broadcast News Items Using Language Models of Concepts",

abstract = "Current video search systems commonly return video shots as results. We believe that users may better relate to longer, semantic video units and propose a retrieval framework for news story items, which consist of multiple shots. The framework is divided into two parts: (1) A concept based language model which ranks news items with known occurrences of semantic concepts by the probability that an important concept is produced from the concept distribution of the news item and (2) a probabilistic model of the uncertain presence, or risk, of these concepts. In this paper we use a method to evaluate the performance of story retrieval, based on the TRECVID shot-based retrieval groundtruth. Our experiments on the TRECVID 2005 collection show a significant performance improvement against four standard methods.",

keywords = "IR-71235, EWI-17850, METIS-270807",

author = "Robin Aly and Aiden Doherty and Djoerd Hiemstra and Alan Smeaton",

note = "10.1007/978-3-642-12275-0\_23 ; 32nd European Conference on Information Retrieval, ECIR 2010 ; Conference date: 28-03-2010 Through 31-03-2010",

year = "2010",

doi = "10.1007/978-3-642-12275-0\_23",

language = "Undefined",

isbn = "978-3-642-12274-3",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "241--252",

editor = "Cathal Gurrin and Yulan He and Gabriella Kazai and Udo Kruschwitz and Suzanne Little and Thomas Roelleke and Stefan R{\"u}ger and \{van Rijsbergen\}, Keith",

booktitle = "Proceedings of the 32nd European Conference on IR Research (ECIR 2010)",

address = "Germany",

}

Aly, R, Doherty, A, Hiemstra, D & Smeaton, A 2010, Beyond Shot Retrieval: Searching for Broadcast News Items Using Language Models of Concepts. in C Gurrin, Y He, G Kazai, U Kruschwitz, S Little, T Roelleke, S Rüger & K van Rijsbergen (eds), Proceedings of the 32nd European Conference on IR Research (ECIR 2010)., 10.1007/978-3-642-12275-0_23, Lecture Notes in Computer Science, vol. 5993/2010, Springer, London, pp. 241-252, 32nd European Conference on Information Retrieval, ECIR 2010, Milton Keynes, United Kingdom, 28/03/10. https://doi.org/10.1007/978-3-642-12275-0_23

Beyond Shot Retrieval: Searching for Broadcast News Items Using Language Models of Concepts. / Aly, Robin; Doherty, Aiden; Hiemstra, Djoerd et al.
Proceedings of the 32nd European Conference on IR Research (ECIR 2010). ed. / Cathal Gurrin; Yulan He; Gabriella Kazai; Udo Kruschwitz; Suzanne Little; Thomas Roelleke; Stefan Rüger; Keith van Rijsbergen. London: Springer, 2010. p. 241-252 10.1007/978-3-642-12275-0_23 (Lecture Notes in Computer Science; Vol. 5993/2010).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - Beyond Shot Retrieval: Searching for Broadcast News Items Using Language Models of Concepts

AU - Aly, Robin

AU - Doherty, Aiden

AU - Hiemstra, Djoerd

AU - Smeaton, Alan

N1 - Conference code: 32

PY - 2010

Y1 - 2010

N2 - Current video search systems commonly return video shots as results. We believe that users may better relate to longer, semantic video units and propose a retrieval framework for news story items, which consist of multiple shots. The framework is divided into two parts: (1) A concept based language model which ranks news items with known occurrences of semantic concepts by the probability that an important concept is produced from the concept distribution of the news item and (2) a probabilistic model of the uncertain presence, or risk, of these concepts. In this paper we use a method to evaluate the performance of story retrieval, based on the TRECVID shot-based retrieval groundtruth. Our experiments on the TRECVID 2005 collection show a significant performance improvement against four standard methods.

AB - Current video search systems commonly return video shots as results. We believe that users may better relate to longer, semantic video units and propose a retrieval framework for news story items, which consist of multiple shots. The framework is divided into two parts: (1) A concept based language model which ranks news items with known occurrences of semantic concepts by the probability that an important concept is produced from the concept distribution of the news item and (2) a probabilistic model of the uncertain presence, or risk, of these concepts. In this paper we use a method to evaluate the performance of story retrieval, based on the TRECVID shot-based retrieval groundtruth. Our experiments on the TRECVID 2005 collection show a significant performance improvement against four standard methods.

KW - IR-71235

KW - EWI-17850

KW - METIS-270807

U2 - 10.1007/978-3-642-12275-0_23

DO - 10.1007/978-3-642-12275-0_23

M3 - Conference contribution

SN - 978-3-642-12274-3

T3 - Lecture Notes in Computer Science

SP - 241

EP - 252

BT - Proceedings of the 32nd European Conference on IR Research (ECIR 2010)

A2 - Gurrin, Cathal

A2 - He, Yulan

A2 - Kazai, Gabriella

A2 - Kruschwitz, Udo

A2 - Little, Suzanne

A2 - Roelleke, Thomas

A2 - Rüger, Stefan

A2 - van Rijsbergen, Keith

PB - Springer

CY - London

T2 - 32nd European Conference on Information Retrieval, ECIR 2010

Y2 - 28 March 2010 through 31 March 2010

ER -

Aly R, Doherty A, Hiemstra D, Smeaton A. Beyond Shot Retrieval: Searching for Broadcast News Items Using Language Models of Concepts. In Gurrin C, He Y, Kazai G, Kruschwitz U, Little S, Roelleke T, Rüger S, van Rijsbergen K, editors, Proceedings of the 32nd European Conference on IR Research (ECIR 2010). London: Springer. 2010. p. 241-252. 10.1007/978-3-642-12275-0_23. (Lecture Notes in Computer Science). doi: 10.1007/978-3-642-12275-0_23