Combining Document Representations for Known-item Search
P. Ogilvie, and J. Callan. Proceedings of the 26th Annual International ACM SIGIR Conference on Research and Development in Informaion Retrieval, page 143--150. New York, NY, USA, ACM, (2003)
DOI: 10.1145/860435.860463
Abstract
This paper investigates the pre-conditions for successful combination of document representations formed from structural markup for the task of known-item search. As this task is very similar to work in meta-search and data fusion, we adapt several hypotheses from those research areas and investigate them in this context. To investigate these hypotheses, we present a mixture-based language model and also examine many of the current meta-search algorithms. We find that compatible output from systems is important for successful combination of document representations. We also demonstrate that combining low performing document representations can improve performance, but not consistently. We find that the techniques best suited for this task are robust to the inclusion of poorly performing document representations. We also explore the role of variance of results across systems and its impact on the performance of fusion, with the surprising result that the correct documents have higher variance across document representations than highly ranking incorrect documents.
%0 Conference Paper
%1 citeulike:1815232
%A Ogilvie, Paul
%A Callan, Jamie
%B Proceedings of the 26th Annual International ACM SIGIR Conference on Research and Development in Informaion Retrieval
%C New York, NY, USA
%D 2003
%I ACM
%K social-search
%P 143--150
%R 10.1145/860435.860463
%T Combining Document Representations for Known-item Search
%U http://dx.doi.org/10.1145/860435.860463
%X This paper investigates the pre-conditions for successful combination of document representations formed from structural markup for the task of known-item search. As this task is very similar to work in meta-search and data fusion, we adapt several hypotheses from those research areas and investigate them in this context. To investigate these hypotheses, we present a mixture-based language model and also examine many of the current meta-search algorithms. We find that compatible output from systems is important for successful combination of document representations. We also demonstrate that combining low performing document representations can improve performance, but not consistently. We find that the techniques best suited for this task are robust to the inclusion of poorly performing document representations. We also explore the role of variance of results across systems and its impact on the performance of fusion, with the surprising result that the correct documents have higher variance across document representations than highly ranking incorrect documents.
%@ 1-58113-646-3
@inproceedings{citeulike:1815232,
abstract = {{This paper investigates the pre-conditions for successful combination of document representations formed from structural markup for the task of known-item search. As this task is very similar to work in meta-search and data fusion, we adapt several hypotheses from those research areas and investigate them in this context. To investigate these hypotheses, we present a mixture-based language model and also examine many of the current meta-search algorithms. We find that compatible output from systems is important for successful combination of document representations. We also demonstrate that combining low performing document representations can improve performance, but not consistently. We find that the techniques best suited for this task are robust to the inclusion of poorly performing document representations. We also explore the role of variance of results across systems and its impact on the performance of fusion, with the surprising result that the correct documents have higher variance across document representations than highly ranking incorrect documents.}},
added-at = {2017-11-15T17:02:25.000+0100},
address = {New York, NY, USA},
author = {Ogilvie, Paul and Callan, Jamie},
biburl = {https://www.bibsonomy.org/bibtex/2930ef054759f611fd6c542fa232d215e/brusilovsky},
booktitle = {Proceedings of the 26th Annual International ACM SIGIR Conference on Research and Development in Informaion Retrieval},
citeulike-article-id = {1815232},
citeulike-linkout-0 = {http://portal.acm.org/citation.cfm?id=860463},
citeulike-linkout-1 = {http://dx.doi.org/10.1145/860435.860463},
doi = {10.1145/860435.860463},
interhash = {ce17fc024612e318e3e0e70df4901862},
intrahash = {930ef054759f611fd6c542fa232d215e},
isbn = {1-58113-646-3},
keywords = {social-search},
location = {Toronto, Canada},
pages = {143--150},
posted-at = {2016-05-17 14:44:36},
priority = {2},
publisher = {ACM},
series = {SIGIR '03},
timestamp = {2017-11-15T17:02:25.000+0100},
title = {{Combining Document Representations for Known-item Search}},
url = {http://dx.doi.org/10.1145/860435.860463},
year = 2003
}