Ranking Relations Using Analogies in Biological and Information Networks

Ranking Relations Using Analogies in Biological and Information Networks

DSpace/Manakin Repository

Ranking Relations Using Analogies in Biological and Information Networks

dc.contributor.author	Airoldi, Edoardo Maria
dc.contributor.author	Silva, Ricardo
dc.contributor.author	Ghahramani, Zoubin
dc.contributor.author	Katherine Heller
dc.date.accessioned	2011-01-05T16:12:30Z
dc.date.issued	2010
dc.identifier.citation	Silva, Ricardo, Katherine Heller, Zoubin Ghahramani, and Edoardo M. Airoldi. 2010. Ranking relations using analogies in biological information networks. Annals of Applied Statistics 4(2): 615-644.	en_US
dc.identifier.issn	1932-6157	en_US
dc.identifier.uri	http://nrs.harvard.edu/urn-3:HUL.InstRepos:4636007
dc.description.abstract	Analogical reasoning depends fundamentally on the ability to learn and generalize about relations between objects. We develop an approach to rela- tional learning which, given a set of pairs of objects S = {A[super](1) : B[super](1) , A[super](2) : B[super](2),...,A[super](N) :B[super](N)}, measures how well other pairs A:B fit in with the set S. Our work addresses the following question: is the relation between objects A and B analogous to those relations found in S? Such questions are particularly relevant in information retrieval, where an investigator might want to search for analogous pairs of objects that match the query set of interest. There are many ways in which objects can be related, making the task of measuring analogies very challenging. Our approach combines a similarity measure on function spaces with Bayesian analysis to produce a ranking. It requires data containing features of the objects of interest and a link matrix specifying which relationships exist; no further attributes of such relationships are necessary. We illustrate the potential of our method on text analysis and information networks. An application on discovering functional interactions between pairs of proteins is discussed in detail, where we show that our approach can work in practice even if a small set of protein pairs is provided.	en_US
dc.description.sponsorship	Statistics	en_US
dc.language.iso	en_US	en_US
dc.publisher	Institute of Mathematical Statistics	en_US
dc.relation.isversionof	doi:10.1214/09-AOAS321	en_US
dc.relation.hasversion	http://arxiv.org/abs/0912.5193/	en_US
dash.license	LAA
dc.subject	network analysis	en_US
dc.subject	Bayesian inference	en_US
dc.subject	variational approximation	en_US
dc.subject	ranking	en_US
dc.subject	information retrieval	en_US
dc.subject	data integration	en_US
dc.subject	Saccharomyces cerevisiae	en_US
dc.title	Ranking Relations Using Analogies in Biological and Information Networks	en_US
dc.type	Journal Article	en_US
dc.description.version	Version of Record	en_US
dc.relation.journal	Annals of Applied Statistics	en_US
dash.depositing.author	Airoldi, Edoardo Maria
dc.date.available	2011-01-05T16:12:30Z