Ranked accuracy and unstructured distributed search

Sami Richardson; Ingemar J Cox

doi:10.1007/978-3-642-36973-5_15

Ranked accuracy and unstructured distributed search

2 Citations (Scopus)

Abstract

Non-uniformly distributing documents in an unstructured peer-to-peer (P2P) network has been shown to improve both the expected search length and search accuracy, where accuracy is defined as the size of the intersection of the documents retrieved by a constrained, probabilistic search and the documents that would have been retrieved by an exhaustive search, normalized by the size of the latter. However neither metric considers the relative ranking of the documents in the retrieved sets. We therefore introduce a new performance metric, rank-accuracy, that is a rank weighted score of the top-k documents retrieved. By replicating documents across nodes based on their retrieval rate (a function of query frequency), and rank, we show that average rank-accuracy can be improved. The practical performance of rank-aware search is demonstrated using a simulated network of 10,000 nodes and queries drawn from a Yahoo! web search log.

Original language	English
Title of host publication	Advances in information retrieval : 5th European Conference on IR Research, ECIR 2013, Moscow, Russia, March 24-27, 2013. Proceedings
Number of pages	12
Publisher	Springer
Publication date	2013
Pages	171-182
Chapter	15
ISBN (Print)	978-3-642-36972-8
ISBN (Electronic)	978-3-642-36973-5
DOIs	https://doi.org/10.1007/978-3-642-36973-5_15
Publication status	Published - 2013
Externally published	Yes
Event	European Conference on IR Research, ECIR 2013 - Moscow, Russian Federation Duration: 24 Mar 2013 → 27 Mar 2013 Conference number: 35

Conference

Conference	European Conference on IR Research, ECIR 2013
Number	35
Country/Territory	Russian Federation
City	Moscow
Period	24/03/2013 → 27/03/2013

Series	Lecture notes in computer science
Volume	7814
ISSN	0302-9743

Access to Document

10.1007/978-3-642-36973-5_15

Cite this

Ranked accuracy and unstructured distributed search. / Richardson, Sami; Cox, Ingemar J.

Advances in information retrieval: 5th European Conference on IR Research, ECIR 2013, Moscow, Russia, March 24-27, 2013. Proceedings. Springer, 2013. p. 171-182 (Lecture notes in computer science, Vol. 7814).

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

Richardson, S & Cox, IJ 2013, Ranked accuracy and unstructured distributed search. in Advances in information retrieval: 5th European Conference on IR Research, ECIR 2013, Moscow, Russia, March 24-27, 2013. Proceedings. Springer, Lecture notes in computer science, vol. 7814, pp. 171-182, European Conference on IR Research, ECIR 2013, Moscow, Russian Federation, 24/03/2013. https://doi.org/10.1007/978-3-642-36973-5_15

@inproceedings{d29ff506e05044adbffce86b3c84e3be,

title = "Ranked accuracy and unstructured distributed search",

abstract = "Non-uniformly distributing documents in an unstructured peer-to-peer (P2P) network has been shown to improve both the expected search length and search accuracy, where accuracy is defined as the size of the intersection of the documents retrieved by a constrained, probabilistic search and the documents that would have been retrieved by an exhaustive search, normalized by the size of the latter. However neither metric considers the relative ranking of the documents in the retrieved sets. We therefore introduce a new performance metric, rank-accuracy, that is a rank weighted score of the top-k documents retrieved. By replicating documents across nodes based on their retrieval rate (a function of query frequency), and rank, we show that average rank-accuracy can be improved. The practical performance of rank-aware search is demonstrated using a simulated network of 10,000 nodes and queries drawn from a Yahoo! web search log.",

author = "Sami Richardson and Cox, {Ingemar J}",

year = "2013",

doi = "10.1007/978-3-642-36973-5_15",

language = "English",

isbn = "978-3-642-36972-8",

series = "Lecture notes in computer science",

publisher = "Springer",

pages = "171--182",

booktitle = "Advances in information retrieval",

note = "European Conference on IR Research, ECIR 2013 ; Conference date: 24-03-2013 Through 27-03-2013",

}

TY - GEN

T1 - Ranked accuracy and unstructured distributed search

AU - Richardson, Sami

AU - Cox, Ingemar J

N1 - Conference code: 35

PY - 2013

Y1 - 2013

N2 - Non-uniformly distributing documents in an unstructured peer-to-peer (P2P) network has been shown to improve both the expected search length and search accuracy, where accuracy is defined as the size of the intersection of the documents retrieved by a constrained, probabilistic search and the documents that would have been retrieved by an exhaustive search, normalized by the size of the latter. However neither metric considers the relative ranking of the documents in the retrieved sets. We therefore introduce a new performance metric, rank-accuracy, that is a rank weighted score of the top-k documents retrieved. By replicating documents across nodes based on their retrieval rate (a function of query frequency), and rank, we show that average rank-accuracy can be improved. The practical performance of rank-aware search is demonstrated using a simulated network of 10,000 nodes and queries drawn from a Yahoo! web search log.

AB - Non-uniformly distributing documents in an unstructured peer-to-peer (P2P) network has been shown to improve both the expected search length and search accuracy, where accuracy is defined as the size of the intersection of the documents retrieved by a constrained, probabilistic search and the documents that would have been retrieved by an exhaustive search, normalized by the size of the latter. However neither metric considers the relative ranking of the documents in the retrieved sets. We therefore introduce a new performance metric, rank-accuracy, that is a rank weighted score of the top-k documents retrieved. By replicating documents across nodes based on their retrieval rate (a function of query frequency), and rank, we show that average rank-accuracy can be improved. The practical performance of rank-aware search is demonstrated using a simulated network of 10,000 nodes and queries drawn from a Yahoo! web search log.

U2 - 10.1007/978-3-642-36973-5_15

DO - 10.1007/978-3-642-36973-5_15

M3 - Article in proceedings

SN - 978-3-642-36972-8

T3 - Lecture notes in computer science

SP - 171

EP - 182

BT - Advances in information retrieval

PB - Springer

T2 - European Conference on IR Research, ECIR 2013

Y2 - 24 March 2013 through 27 March 2013

ER -

Ranked accuracy and unstructured distributed search

Abstract

Conference

Access to Document

Fingerprint

Cite this