W. Bruce Croft (auth.), Mohand Boughanem, Catherine Berrut,'s Advances in Information Retrieval: 31th European Conference PDF

By W. Bruce Croft (auth.), Mohand Boughanem, Catherine Berrut, Josiane Mothe, Chantal Soule-Dupuy (eds.)

ISBN-10: 3642009573

ISBN-13: 9783642009570

ISBN-10: 3642009581

ISBN-13: 9783642009587

This publication constitutes the refereed complaints of the thirtieth annual ecu convention on info Retrieval examine, ECIR 2009, held in Toulouse, France in April 2009.

The forty two revised complete papers and 18 revised brief papers offered including the abstracts of three invited lectures and 25 poster papers have been conscientiously reviewed and chosen from 188 submissions. The papers are prepared in topical sections on retrieval version, collaborative IR / filtering, studying, multimedia - metadata, specialist seek - ads, assessment, opinion detection, net IR, illustration, clustering / categorization in addition to dispensed IR.

Show description

Read or Download Advances in Information Retrieval: 31th European Conference on IR Research, ECIR 2009, Toulouse, France, April 6-9, 2009. Proceedings PDF

Best european books

A Continent Moving West?: EU Enlargement and Labour by Richard Black, Godfried Engbersen, Marek Okólski, Cristina PDF

  A Continent relocating West? argues that the conceptualization of migration as a one-way or long term technique is changing into more and more extensive of the mark. particularly, east-west exertions migration in Europe, in universal maybe with different flows in and from different components of the area, is varied, fluid, and inspired by means of the dynamics of neighborhood and sector-specific hard work markets and migration-related political laws.

Download e-book for kindle: New Modes of Governance in Europe: Governing in the Shadow by Adrienne Heritier, Martin Rhodes

In line with the learn of the EU-6th framework funded learn consortium on ‘New Modes of Governance within the eu Union’, this quantity explores the roots, execution and functions of latest types of governance and evaluates their luck.

Erik Kristensen (auth.), Gerd Liebezeit, Sabine Dittmann,'s Life at Interfaces and Under Extreme Conditions: Proceedings PDF

Interfaces among media, even if air-water or sediment-water interfaces or organisms themselves, pose massive difficulties to marine organisms trying to stay at those obstacles. within the current quantity, a couple of authors deal with numerous points of those subject matters. destinations lower than scrutiny variety from intertidal components to the deep sea, whereas either macro-and meiofaunal organisms are investigated.

Additional info for Advances in Information Retrieval: 31th European Conference on IR Research, ECIR 2009, Toulouse, France, April 6-9, 2009. Proceedings

Sample text

A study of smoothing methods for language models applied to ad hoc information retrieval. In: Proc. of SIGIR 2001, pp. 334–342 (2001) 19. : A risk minimization framework for information retrieval. Inf. Process. Manage. 42(1), 31–55 (2006) 20. : Risky business: Modeling and exploiting uncertainty in information retrieval. Technical Report. University College London (2008) Appendix A θt Since f (θt ) = ln 1−θ is infinitely differentiable in the neighborhood of the mean t of θt , the mean of f (θt ) can be approximated as the mean of a Taylor series as: E[f (θt )] 1 = E[f (θ¯t )] + E[(θt − θ¯t )f (θ¯t )] + E[ (θt − θ¯t )2 f (θ¯t )] + · · · 2 1 = f (θ¯t ) + 0 + f (θ¯t )V ar(θt ) + · · · 2 f (θ¯t ) θ¯t 2θ¯t − 1 V ar(θt ) = ln + V ar(θt ) ≈ f (θ¯t ) + 2 1 − θ¯t 2θ¯t2 (1 − θ¯t )2 (11) f (θt ) can be approximated by a first order Taylor series as f (θt ) ≈ f (θ¯t ) + (θt − θ¯t )f (θ¯t ).

Finally, we conclude in Section 6. g. maximum likelihood estimation, recent studies have focused on building more accurate language models for documents, including background smoothing based on collection statistics [18], and a Bayesian treatment of the language modelling framework [16]. By considering risk in retrieval, a risk minimization framework was proposed in [19] for ranking documents based on the expected risk of these documents. The framework has been applied to subtopic retrieval for modelling redundancy and novelty in addition to relevance.

Positive and statistically significant improvements are in bold, and in bold and marked with “*”, respectively. 05. (a) LM with Dirichlet smoothing (μ = 2000) vs. 1) vs. 4% parameter, b, controls how much risk we are prepared to take when ranking documents, and the effect this has on the result set. , risk-aversion (conservative ranking) we have a much greater chance that at least one document will be relevant, but the chance that many of the documents will be relevant is diminished. Conversely, for a risk-loving (aggressive ranking), we have a much greater chance that many of the documents will be relevant, but at the expense that some searches produce no relevant documents.

Download PDF sample

Advances in Information Retrieval: 31th European Conference on IR Research, ECIR 2009, Toulouse, France, April 6-9, 2009. Proceedings by W. Bruce Croft (auth.), Mohand Boughanem, Catherine Berrut, Josiane Mothe, Chantal Soule-Dupuy (eds.)


by James
4.4

Rated 4.63 of 5 – based on 15 votes