CiteULike: Author Callan

Retrieval and feedback models for blog feed search

Jonathan Elsas — 2008-07-31T06:07:20-00:00

(2008), pp. 347-354.

The impact of history length on personalized search

Yangbo Zhu — 2008-07-22T08:09:23-00:00

(2008), pp. 715-716.

Specificity of T cells in synovial fluid: high frequencies of CD8(+) T cells that are specific for certain viral epitopes.

LC Tan — 2008-07-26T02:45:36-00:00

Arthritis research, Vol. 2, No. 2. (2000), pp. 154-164.

INTRODUCTION: Epstein-Barr virus (EBV) is transmitted orally, replicates in the oropharynx and establishes life-long latency in human B lymphocytes. T-cell responses to latent and lytic/replicative cycle proteins are readily detectable in peripheral blood from healthy EBV-seropositive individuals. EBV has also been detected within synovial tissue, and T-cell responses to EBV lytic proteins have been reported in synovial fluid from a patient with rheumatoid arthritis (RA). This raises the question regarding whether T cells specific for certain viruses might be present at high frequencies within synovial fluid and whether such T cells might be activated or able to secrete cytokines. If so, they might play a 'bystander' role in the pathogenesis of inflammatory joint disease. OBJECTIVES: To quantify and characterize T cells that are specific for epitopes from EBV, cytomegalovirus (CMV) and influenza in peripheral blood and synovial fluid from patients with arthritis. Methods: Peripheral blood mononuclear cells (PBMCs) and synovial fluid mononuclear cells (SFMCs) were obtained from patients with inflammatory arthritis (including those with RA, osteoarthritis, psoriatic arthritis and reactive arthritis). Samples from human leucocyte antigen (HLA)-A2-positive donors were stained with fluorescent-labelled tetramers of HLA-A2 complexed with the GLCTLVAML peptide epitope from the EBV lytic cycle protein BMLF1, the GILGFVFTL peptide epitope from the influenza A matrix protein, or the NLVPMVATV epitope from the CMV pp65 protein. Samples from HLA-B8-positive donors were stained with fluorescent-labelled tetramers of HLA-B8 complexed with the RAKFKQLL peptide epitope from the EBV lytic protein BZLF1 or the FLRGRAYGL peptide epitope from the EBV latent protein EBNA3A. All samples were costained with an antibody specific for CD8. CD4+ T cells were not analyzed. Selected samples were costained with antibodies specific for cell-surface glycoproteins, in order to determine the phenotype of the T cells within the joint and the periphery. Functional assays to detect release of IFN- or tumour necrosis factor (TNF)- were also performed on some samples. RESULTS: The first group of 15 patients included 10 patients with RA, one patient with reactive arthritis, one patient with psoriatic arthritis and three patients with osteoarthritis. Of these, 11 were HLA-A2 positive and five were HLA-B8 positive. We used HLA-peptide tetrameric complexes to analyze the frequency of EBV-specific T cells in PBMCs and SFMCs (Figs 1 and 2). Clear enrichment of CD8+ T cells specific for epitopes from the EBV lytic cycle proteins was seen within synovial fluid from almost all donors studied, including patients with psoriatic arthritis and osteoarthritis and those with RA. In donor RhA6, 9.5% of CD8+ SFMCs were specific for the HLA-A2 restricted GLCTLVAML epitope, compared with 0.5% of CD8+ PBMCs. Likewise in a donor with osteoarthritis (NR4), 15.5% of CD8+ SFMCs were specific for the HLA-B8-restricted RAKFKQLL epitope, compared with 0.4% of CD8+ PBMCs. In contrast, we did not find enrichment of T cells specific for the HLA-B8-restricted FLRGRAYGL epitope (from the latent protein EBNA3A) within SFMCs compared with PBMCs in any donors. In selected individuals we performed ELISpot assays to detect IFN- secreted by SFMCs and PBMCs after a short incubation in vitro with peptide epitopes from EBV lytic proteins. These assays confirmed enrichment of T cells specific for epitopes from EBV lytic proteins within synovial fluid and showed that subpopulations of these cells were able to secrete proinflammatory cytokines after short-term stimulation. We used a HLA-A2/GILGFVFTL tetramer to stain PBMCs and SFMCs from six HLA-A2-positive patients. The proportion of T cells specific for this influenza epitope was low (<0.2%) in all donors studied, and we did not find any enrichment within SFMCs. We had access to SFMCs only from a second group of four HLA-A2-positive patients with RA. A tetramer of HLA-A2 complexed to the NLVPMVATV epitope from the CMV pp65 protein reacted with subpopulations of CD8+ SFMCs in all four donors, with frequencies of 0.2, 0.5, 2.3 and 13.9%. SFMCs from all four donors secreted TNF after short-term incubation with COS cells transfected with HLA-A2 and pp65 complementary DNA. We analyzed the phenotype of virus-specific cells within PBMCs and SFMCs in three donors. The SFMC virus-specific T cells were more highly activated than those in PBMCs, as evidenced by expression of high levels of CD69 and HLA-DR. A greater proportion of SFMCs were CD38+, CD62L low, CD45RO bright, CD45RA dim, CD57+ and CD28- when compared with PBMCs. Discussion:This work shows that T cells specific for certain epitopes from viral proteins are present at very high frequencies (up to 15.5% of CD8+ T cells) within SFMCs taken from patients with inflammatory joint disease. This enrichment does not reflect a generalized enrichment for the 'memory pool' of T cells; we did not find enrichment of T cells specific for the GILGFVFTL epitope from influenza A or for the FLRGRAYGL epitope from the EBV latent protein EBNA3A, whereas we found clear enrichment of T cells specific for the GLCTLVAML epitope from the EBV lytic protein BMLF1 and for the RAKFKQLL epitope from the EBV lytic protein BZLF1. The enrichment might reflect preferential recruitment of subpopulations of virus-specific T cells, perhaps based on expression of selectins, chemokine receptors or integrins. Alternatively, T cells specific for certain viral epitopes may be stimulated to proliferate within the joint, by viral antigens themselves or by cross-reactive self-antigens. Finally, it is theoretically possible that subpopulations of T cells within the joint are preferentially protected from apoptotic cell death. Whatever the explanation, the virus-specific T cells are present at high frequency, are activated and are able to secrete proinflammatory cytokines. They could potentially interact with synoviocytes and contribute to the maintenance of inflammation within joints in many different forms of inflammatory arthritis.

Experiments Using the Lemur Toolkit

Paul Ogilvie — 2008-07-16T10:07:50-00:00

(2001)

An auditory-feedback-based neural network model of speech production that is robust to developmental changes in the size and shape of the articulatory system.

DE Callan — 2008-07-12T20:23:19-00:00

Journal of speech, language, and hearing research : JSLHR, Vol. 43, No. 3. (June 2000), pp. 721-736.

The purpose of this article is to demonstrate that self-produced auditory feedback is sufficient to train a mapping between auditory target space and articulator space under conditions in which the structures of speech production are undergoing considerable developmental restructuring. One challenge for competing theories that propose invariant constriction targets is that it is unclear what teaching signal could specify constriction location and degree so that a mapping between constriction target space and articulator space can be learned. It is predicted that a model trained by auditory feedback will accomplish speech goals, in auditory target space, by continuously learning to use different articulator configurations to adapt to the changing acoustic properties of the vocal tract during development. The Maeda articulatory synthesis part of the DIVA neural network model (Guenther et al., 1998) was modified to reflect the development of the vocal tract by using measurements taken from MR images of children. After training, the model was able to maintain the 11 English vowel targets in auditory planning space, utilizing varying articulator configurations, despite morphological changes that occur during development. The vocal-tract constriction pattern (derived from the vocal-tract area function) as well as the formant values varied during the course of development in correspondence with morphological changes in the structures involved with speech production. Despite changes in the acoustical properties of the vocal tract that occur during the course of development, the model was able to demonstrate motor-equivalent speech production under lip-restriction conditions. The model accomplished this in a self-organizing manner even though there was no prior experience with lip restriction during training.

Collaborative filtering with decoupled models for preferences and ratings

Rong Jin — 2008-07-12T15:30:41-00:00

(2003), pp. 309-316.

Meeting of the MINDS: an information retrieval research agenda

Jamie Callan — 2008-05-06T15:12:56-00:00

SIGIR Forum, Vol. 41, No. 2. (December 2007), pp. 25-34.

We Do It, but They Don't: Multiple Categorizations and Work Team Communication

Grice — 2006-09-25T04:32:13-00:00

Journal of Applied Communication Research, Vol. 34, No. 4. (November 2006), pp. 331-348.

Document filtering with inference networks

Jamie Callan — 2008-04-24T14:24:00-00:00

(1996), pp. 262-269.

Giving speech a hand: Gesture modulates activity in auditory cortex during speech perception.

Amy L Hubbard — 2008-04-22T01:41:52-00:00

Human brain mapping (15 April 2008)

Viewing hand gestures during face-to-face communication affects speech perception and comprehension. Despite the visible role played by gesture in social interactions, relatively little is known about how the brain integrates hand gestures with co-occurring speech. Here we used functional magnetic resonance imaging (fMRI) and an ecologically valid paradigm to investigate how beat gesture-a fundamental type of hand gesture that marks speech prosody-might impact speech perception at the neural level. Subjects underwent fMRI while listening to spontaneously-produced speech accompanied by beat gesture, nonsense hand movement, or a still body; as additional control conditions, subjects also viewed beat gesture, nonsense hand movement, or a still body all presented without speech. Validating behavioral evidence that gesture affects speech perception, bilateral nonprimary auditory cortex showed greater activity when speech was accompanied by beat gesture than when speech was presented alone. Further, the left superior temporal gyrus/sulcus showed stronger activity when speech was accompanied by beat gesture than when speech was accompanied by nonsense hand movement. Finally, the right planum temporale was identified as a putative multisensory integration site for beat gesture and speech (i.e., here activity in response to speech accompanied by beat gesture was greater than the summed responses to speech alone and beat gesture alone), indicating that this area may be pivotally involved in synthesizing the rhythmic aspects of both speech and gesture. Taken together, these findings suggest a common neural substrate for processing speech and gesture, likely reflecting their joint communicative role in social interactions. Hum Brain Mapp, 2008. (c) 2008 Wiley-Liss, Inc.

Parameter Estimation for a Simple Hierarchical Generative Model for XML Retrieval

Paul Ogilvie — 2008-01-20T21:56:01-00:00

Advances in XML Information Retrieval and Evaluation (2006), pp. 211-224.

This paper explores the possibility of using a modified Expectation-Maximization algorithm to estimate parameters for a simple hierarchical generative model for XML retrieval. The generative model for an XML element is estimated by linearly interpolating statistical language models estimated from the text of the element, the parent element, the document element, and its children elements. We heuristically modify EM to allow the incorporation of negative examples, then attempt to maximize the likelihood of the relevant components while minimizing the likelihood of non-relevant components found in training data. The technique for incorporation of negative examples provide an effective algorithm to estimate the parameters in the linear combination mentioned. Some experiments are presented on the CO.Thorough task that support these claims.

Precise physical models of protein-DNA interaction from high-throughput data

Justin Kinney — 2007-01-10T08:58:33-00:00

PNAS, Vol. 104, No. 2. (9 January 2007), pp. 501-506.

A cell's ability to regulate gene transcription depends in large part on the energy with which transcription factors (TFs) bind their DNA regulatory sites. Obtaining accurate models of this binding energy is therefore an important goal for quantitative biology. In this article, we present a principled likelihood-based approach for inferring physical models of TF-DNA binding energy from the data produced by modern high-throughput binding assays. Central to our analysis is the ability to assess the relative likelihood of different model parameters given experimental observations. We take a unique approach to this problem and show how to compute likelihood without any explicit assumptions about the noise that inevitably corrupts such measurements. Sampling possible choices for model parameters according to this likelihood function, we can then make probabilistic predictions for the identities of binding sites and their physical binding energies. Applying this procedure to previously published data on the Saccharomyces cerevisiae TF Abf1p, we find models of TF binding whose parameters are determined with remarkable precision. Evidence for the accuracy of these models is provided by an astonishing level of phylogenetic conservation in the predicted energies of putative binding sites. Results from in vivo and in vitro experiments also provide highly consistent characterizations of Abf1p, a result that contrasts with a previous analysis of the same data. 10.1073/pnas.0609908104

User modeling for full-text federated search in peer-to-peer networks

Jie Lu — 2008-01-23T20:02:07-00:00

(2006), pp. 332-339.

User modeling for information retrieval has mostly been studied to improve the effectiveness of information access in centralized repositories. In this paper we explore user modeling in the context of full-text federated search in peer-to-peer networks. Our approach models a user's persistent, long-term interests based on past queries, and uses the model to improve search efficiency for future queries that represent interests similar to past queries. Our approach also enables queries representing a user's transient, ad-hoc interests to be automatically recognized so that search for these queries can rely on a relatively large search radius to avoid sacrificing effectiveness for efficiency. Experimental results demonstrate that our approach can significantly improve the efficiency of full-text federated search without degrading its accuracy. Furthermore, the proposed approach does not require a large amount of training data, and is robust to a range of parameter values.

Structured retrieval for question answering

Matthew Bilotti — 2008-03-16T08:47:52-00:00

SIGIR (2007), pp. 351-358.

Distributed information retrieval

J Callan — 2007-07-31T05:47:29-00:00

(2000)

A multi-database model of distributed information retrieval is presented, in which people are assumed to have access to many searchable text databases. In such an environment, full-text information retrieval consists of discovering database contents, ranking databases by their expected ability to satisfy the query, searching a small number of databases, and merging results returned by different databases. This paper presents algorithms for each task. It also discusses how to reorganize...

The Effectiveness of Query Expansion for Distributed Information Retrieval

Paul Ogilvie — 2007-07-31T02:21:19-00:00

(2001), pp. 183-190.

Query expansion has been shown effective for both single database retrieval and for distributed information retrieval where complete collection information is available. One might expect that query expansion would then work for distributed information retrieval when complete collection information is not available. However, this does not appear to be the case. When using local context analysis for query expansion in distributed retrieval with partial information, the most significant reason...

The Effects of Query-Based Sampling on Automatic Database Selection Algorithms

J Callan — 2008-02-21T15:34:27-00:00

(2000)

Database selection algorithms need to know the subject areas covered by each text database, but this metadata can be difficult to acquire in multi-party environments, such as the Internet, where each party has different interests and capabilities. Query-based sampling is a relatively new technique in which metadata is inferred by interacting with each text database and observing the outcomes. Query-based sampling has been proposed as a solution to the problem of discovering the contents of...

Effective and Efficient Automatic Database Selection

James French — 2008-02-21T15:34:25-00:00

No. CS-99-08. (FebruaryFebruary, 1999)

We examine a class of database selection algorithms that require only document frequency information. The CORI algorithm is an instance of this class of algorithms. In previous work, we showed that CORI is more effective than gGlOSS when evaluated against a relevance-based standard. In this paper, we introduce a family of other algorithms in this class and examine components of these algorithms and of the CORI algorithm to begin identifying the factors responsible for their performance. We...

Passage-level evidence in document retrieval

James Callan — 2008-02-19T05:57:30-00:00

(1994), pp. 302-310.

Novelty and redundancy detection in adaptive filtering

Yi Zhang — 2008-01-27T19:27:51-00:00

(2002), pp. 81-88.

Near-duplicate detection by instance-level constrained clustering

Hui Yang — 2008-01-27T19:17:59-00:00

(2006), pp. 421-428.

Hierarchical Language Models for XML Component Retrieval

Paul Ogilvie — 2008-01-20T21:51:41-00:00

Advances in XML Information Retrieval (2005), pp. 224-237.

Experiments using hierarchical language models for XML component retrieval are presented in this paper. The role of context is investigated through incorporation of the parent’s model. We find that context can improve the effectiveness of finding relevant components slightly. Additionally, biasing the results toward long components through the use of component priors improves exhaustivity but harms specificity, so care must be taken to find an appropriate trade-off.

Modeling Annotated Data

DM Blei — 2007-02-05T15:53:46-00:00

SIGIR Forum, No. SPEC. ISS.. (2003), pp. 127-134.

We consider the problem of modeling annotated data-data with multiple types where the instance of one type (such as a caption) serves as a description of the other type (such as an image). We describe three hierarchical probabilistic mixture models which aim to describe such data, culminating in correspondence latent Dirichlet allocation, a latent variable model that is effective at modeling the joint distribution of both types and the conditional distribution of the annotation given the primary type. We conduct experiments on the Corel database of images and captions, assessing performance in terms of held-out likelihood, automatic annotation, and text-based image retrieval. Sponsors: ACM/SIGIR

Query expansion using random walk models

Kevyn Collins-Thompson — 2008-01-02T16:31:14-00:00

(2005), pp. 704-711.

INQUERY System Overview

John Broglio — 2007-10-07T02:55:31-00:00

(1994), pp. 47-67.

such as words, phrases, paragraphs, or manually assigned keywords) and different versions of the query (such as natural language and Boolean) can be combined in a consistent probabilistic framework. This type of "data fusion" has been known to be effective in the information retrieval context for a number of years, and was one of the primary motivations for developing the inference net approach. Another feature of the inference net approach is the ability to capture complex structure in the...

Passage-Level Evidence in Document Retrieval

James Callan — 2006-12-08T06:15:17-00:00

(1994), pp. 302-310.

Estimation and use of uncertainty in pseudo-relevance feedback

Kevyn Collins-Thompson — 2007-09-13T20:28:35-00:00

(2007), pp. 303-310.

Relevant document distribution estimation method for resource selection

Luo Si — 2007-03-20T03:32:37-00:00

(2003), pp. 298-305.

Implications of family-friendly policies for organizational culture: findings from two case studies

Samantha Callan — 2007-11-24T20:12:19-00:00

Work Employment Society, Vol. 21, No. 4. (1 December 2007), pp. 673-691.

Formal policies intended to enable employees to meet family commitments may be important indicators of an organization's intent, but they do not guarantee that the informal culture is supportive of employees' families or their attempts to manage occasionally conflicting priorities (Lewis, 1997; Lewis and Lewis, 1996). Two case studies were conducted to identify salient aspects of the culture of two organizations and the extent to which changes in culture result from the implementation of family-friendly policies.The wider issue of the ease with which purposive cultural change or organizational learning may be engendered to ameliorate employees' worklife balance is also considered. 10.1177/0950017007082876

String loop corrections to beta functions

CG Callan — 2007-11-22T05:05:32-00:00

Nuclear Physics B, Vol. 288 (1987), pp. 525-550.

We study the problem of finding the beta functions, and the associated spacetime effective action, for interacting open and closed strings propagating in background fields. String loop divergences play a crucial role in this problem. Cancelling them against sigma model divergences gives a consistent set of loop-corrected beta functions, which can be derived from a simple generalization of the string-tree-level effective action. This suggests the existence of new string theories which are conformally invariant only after all world sheets have been summed.

Bijections from Dyck paths to 321-avoiding permutations revisited

David Callan — 2007-11-21T05:01:32-00:00

(16 Nov 2007)

There are (at least) three bijections from Dyck paths to 321-avoiding permutations in the literature, due to Billey-Jockusch-Stanley, Krattenthaler, and Mansour-Deng-Du. How different are they? Denoting them B,K,M respectively, we show that M = B \circ L = K \circ L' where L is the classical Kreweras-Lalanne involution on Dyck paths and L', also an involution, is a sort of derivative of L. Thus K^-1 \circ B, a measure of the difference between B and K, is the product of involutions L' \circ L and turns out to be a very curious bijection: as a permutation on Dyck n-paths it is an nth root of the "reverse path" involution. The proof of this fact boils down to a geometric argument involving pairs of nonintersecting lattice paths.

Novelty and redundancy detection in adaptive filtering

Yi Zhang — 2007-11-13T11:55:39-00:00

(2002), pp. 81-88.

Next steps in near-duplicate detection for eRulemaking

Hui Yang — 2007-11-10T09:31:46-00:00

(2006), pp. 239-248.

Combining document representations for known-item search

Paul Ogilvie — 2007-10-24T12:17:36-00:00

(2003), pp. 143-150.

Evolutionary comparisons suggest many novel cAMP response protein binding sites in Escherichia coli.

CT Brown — 2007-10-16T21:09:28-00:00

Proc Natl Acad Sci U S A, Vol. 101, No. 8. (24 February 2004), pp. 2404-2409.

The cAMP response protein (CRP) is a transcription factor known to regulate many genes in Escherichia coli. Computational studies of transcription factor binding to DNA are usually based on a simple matrix model of sequence-dependent binding energy. For CRP, this model predicts many binding sites that are not known to be functional. If they are indeed spurious, the underlying binding model is called into question. We use a species comparison method to assess the functionality of a population of such predicted CRP sites in E. coli. We compare them with orthologous sites in Salmonella typhimurium identified independently by CLUSTALW alignment, and find a dependence of mutation probability on position in the site. This dependence increases with predicted site binding energy. The positions where mutation is most strongly suppressed are those where mutation would have the biggest effect on predicted binding energy. This finding suggests that many of the novel sites are functional, that the matrix model correctly estimates their binding strength, and that calculated CRP binding strength is the quantity that is conserved between species. The analysis also identifies many new E. coli binding sites and genes likely to be functional for CRP.

The INQUERY Retrieval System

James Callan — 2005-03-26T09:21:40-00:00

(1992), pp. 78-83.

As larger and more heterogeneous text databases become available, information retrieval research will depend on the development of powerful, efficient and flexible retrieval engines. In this paper, we describe a retrieval system (INQUERY) that is based on a probabilistic retrieval model and provides support for sophisticated indexing and complex query formulation. INQUERY has been used successfully with databases containing nearly 400,000 documents. 1 Introduction The increasing interest in...

Searching Distributed Collections with Inference Networks

JP Callan — 2007-08-15T06:16:11-00:00

(1995), pp. 21-28.

The use of information retrieval systems in networked environments raises a new set of issues that have received little attention. These issues include ranking document collections for relevance to a query, selecting the best set of collections from a ranked list, and merging the document rankings that are returned from a set of collections. This paper describes methods of addressing each issue in the inference network model, discusses their implementation in the INQUERY system, and presents...

The SIGIR peer-to-peer information retrieval workshop

Jamie Callan — 2007-10-07T01:34:45-00:00

SIGIR Forum, Vol. 38, No. 2. (December 2004), pp. 37-40.

The path to new medicines

Bénédicte Callan — 2007-09-14T14:40:00-00:00

Nature, Vol. 449, No. 7159. (12 September 2007), pp. 164-165.

Reducing storage costs for federated search of text databases

J Lu — 2007-08-15T06:27:28-00:00

(2003)

In environments containing many text search engines a federated search system provides people with a single point of access. When search engines are managed by independent organizations two key problems are discovering and representing the contents of each text database. Query-based sampling is a recent technique for discovering the contents of uncooperative databases so as to create database resource descriptions that support a variety of necessary capabilities. However, when the documents...

A semi-supervised learning method to merge search engine results

L Si — 2007-07-31T02:21:48-00:00

This article presents a semisupervised learning solution to the result merging problem. The key contribution is the observation that information used to create resource descriptions for resource selection can also be used to create a centralized sample database to guide the normalization of document scores returned by different databases. At retrieval time, the query is sent to the selected databases, which return database-specific document scores, and to a centralized sample database, which...

Collection Selection and Results Merging with Topically Organized U.S. Patents and TREC Data

Leah Larkey — 2007-07-31T02:18:36-00:00

(2000), pp. 282-289.

We investigate three issues in distributed information retrieval, considering both TREC data and U.S. Patents: (1) topical organization of large text collections, (2) collection ranking and selection with topically organized collections (3) results merging, particularly document score normalization, with topically organized collections. We find that it is better to organize collections topically, and that topical collections can be well ranked using either INQUERY's CORI algorithm, or the...

Parents' Views on Social Work Interventions in Child Welfare Cases

Trevor Spratt — 2007-07-24T15:10:02-00:00

Br J Soc Work, Vol. 34, No. 2. (1 March 2004), pp. 199-224.

This article reports findings from the third part of a three-part research project examining the potential for social workers to shift from a child protection to a child welfare orientation in their practice. Whilst social workers in the UK have been encouraged to make such changes, they have been hampered by concerns to manage risk. Findings reported from the earlier parts of the project, indicated that there was potential for a substantial proportion of child protection work to be redesignated as child welfare work, but that where this was achieved in practice, there was evidence of the continued influence of child protection processes as social workers sought to manage the risks inherent in child welfare cases. The study reported here sets out to ascertain the views of parents who were subject to child welfare interventions. The findings indicate that while parents feel apprehension with regard to contact with social workers, in the majority of cases successful relationships are formed. It is argued that social workers display considerable skill in monitoring potential risks whilst engaging with families and that the subtleties involved in such activity are not captured by official measures of governance which concentrate on more abstract indicators of performance. 10.1093/bjsw/bch022

Comparing the performance of database selection algorithms

James French — 2007-05-30T07:46:29-00:00

(1999), pp. 238-245.

Comparing the Performance of Database Selection Algorithms

James French — 2007-05-29T12:27:59-00:00

(1999), pp. 238-245.

We compare the performance of two database selection algorithms reported in the literature. Their performance is compared using a common testbed designed specifically for database selection techniques. The testbed is a decomposition of the TREC/TIPSTER data into 236 subcollections. The databases from our testbed were ranked using both the gGlOSS and CORI techniques and compared to a baseline derived from TREC relevance judgements. We examined the degree to which CORI and gGlOSS approximate this ...

Modeling search engine effectiveness for federated search

Luo Si — 2005-10-04T15:28:08-00:00

(2005), pp. 83-90.

Recent experiments with INQUERY

J Allan — 2007-07-18T23:52:05-00:00

(1995)

this paper focuses on relevant differences to the previously published algorithms. 1 Description of Ad-Hoc Experiments

Exact Solution of a Boundary Conformal Field Theory

Curtis Callan — 2007-07-03T02:51:19-00:00

(25 May 2003)

We study the conformal field theory of a free massless scalar field living on the half line with interactions introduced via a periodic potential at the boundary. An SU(2) current algebra underlies this system and the interacting boundary state is given by a global SU(2) rotation of the left-moving fields in the zero-potential (Neumann) boundary state. As the potential strength varies from zero to infinity, the boundary state interpolates between the Neumann and the Dirichlet values. The full S-matrix for scattering from the boundary, with arbitrary particle production, is explicitly computed. To maintain unitarity, it is necessary to attribute a hidden discrete “soliton” degree of freedom to the boundary. The same unitarity puzzle occurs in the Kondo problem, and we anticipate a similar solution.

Personalisation and recommender systems in digital libraries

Alan Smeaton — 2007-02-15T20:26:40-00:00

International Journal on Digital Libraries, Vol. V5, No. 4. (2005), pp. 299-308.

Automatic Image Annotation and Retrieval using Cross-Media Relevance Models

J Jeon — 2007-02-05T15:51:57-00:00

SIGIR Forum, No. SPEC. ISS.., pp. 119-126.

Libraries have traditionally used manual image annotation for indexing and then later retrieving their image collections. However, manual image annotation is an expensive and labor intensive procedure and hence there has been great interest in coming up with automatic ways to retrieve images based on content. Here, we propose an automatic approach to annotating and retrieving images based on a training set of images. We assume that regions in an image can be described using a small vocabulary of blobs. Blobs are generated from image features using clustering. Given a training set of images with annotations, we show that probabilistic models allow us to predict the probability of generating a word given the blobs in an image. This may be used to automatically annotate and retrieve images given a word as a query. We show that relevance models, allow us to derive these probabilities in a natural way. Experiments show that the annotation performance of this cross-media relevance model is almost six times as good (in terms of mean precision) than a model based on word-blob co-occurrence model and twice as good as a state of the art model derived from machine translation. Our approach shows the usefulness of using formal information retrieval models for the task of image annotation and retrieval. Sponsors: ACM/SIGIR