Network Communication as a Service-Oriented Capability (2009)
William Johnston, Joe Metzger, Michael Collins, Eli Dart, Jim Gagliardi, Chin Guok, ...
Abstract. In widely distributed systems generally, and in
TAG, Dynamic Programming, and the Perceptron for Efficient, Feature-rich Parsing (2009)
Xavier Carreras, Michael Collins, Terry Koo
We describe a parsing approach that makes use of the perceptron algorithm, in conjunction with dynamic programming methods, to recover full constituent-based parse trees. The formalism allows a rich...
Transfer Learning for Image Classification with Sparse Prototype Representations (2009)
Ariadna Quattoni, Michael Collins, Trevor Darrell
To learn a new visual category from few examples, prior knowledge from unlabeled data as well as previous related categories may be useful. We develop a new method for transfer learning which...
Signifying in New Orleans: Notes on a Journey into the Imaginal (2009)
Callaloo - Volume 32, Number 2, Spring 2009
Borga, Angel, Collins, Michael
Marine legislation world-wide has, as a recent final objective, the maintenance of a good environmental or ecological status for marine waters, habitats and resources. The concept of environmental...
Fontán, Almudena, González, Manuel, Wells, Neil, Collins, Michael, Mader, Julien, Ferrer, Luis, ...
The Basque coastal area, in the southeastern Bay of Biscay, can be characterised as being more influenced by land climate and inputs, than other typically ‘open sea’ areas. The influence of...
Simon Aumônier, Michael Collins
in the UK The project was informed and assisted by an Advisory Board set up by the Environment Agency. Membership of the Advisory Board was decided by the Environment Agency on the basis of interest,...
Callaloo - Volume 31, Number 4, Fall 2008
Callaloo - Volume 31, Number 4, Fall 2008
“My Preoccupations are in My DNA”: An interview with Bernardine Evaristo (2009)
Callaloo - Volume 31, Number 4, Fall 2008
“My Generation’s Task”: An Interview with Melvin White (2009)
Callaloo - Volume 31, Number 4, Fall 2008
Globalization and the Group of 77: Three Interviews (2009)
Callaloo - Volume 31, Number 4, Fall 2008
Callaloo - Volume 31, Number 4, Fall 2008
Callaloo - Volume 31, Number 4, Fall 2008
“I’m Interested as A Writer in Less Exalted Persons”: An Interview with Jessica Hagedorn (2009)
Callaloo - Volume 31, Number 4, Fall 2008
The Monterrey Process and Beyond: An Interview with Tim Wall (2009)
Callaloo - Volume 31, Number 4, Fall 2008
Learning Visual Representations using Images with Captions (2008)
Ariadna Quattoni, Michael Collins, Trevor Darrell
Current methods for learning visual categories work well when a large amount of labeled data is available, but can run into severe difficulties when the number of labeled examples is small. When...
Dimensionality Reduction for Speech Recognition Using Neighborhood Components Analysis (2008)
Natasha Singh-miller, Michael Collins, Timothy J. Hazen
Previous work has considered methods for learning projections of high-dimensional acoustic representations to lower dimensional spaces. In this paper we apply the neighborhood components analysis...
Thomas Mulja, Michael Collins, Henry H. Wong, Rifqi Rizal, Tyson Brown, Muharis Zainuddin
ABSTRACT: A multidisciplinary approach led to the discovery of several gold and base metal occurrences during the 1996–1998 exploration campaign in the Takengon tenement of the Aceh magmatic arc,...
2006, ‘Statistical LTAG Parsing (2008)
Libin Shen, Aravind K. Joshi, Rajeev Alur, John Blitzer, Jinying Chen, ...
First and foremost, I would like to thank my advisor Aravind Joshi for his continuous support and guidance in both academic and daily life ever since my first day at Penn. Many thanks to my...
TAG, Dynamic Programming, and the Perceptron for Efficient, Feature-rich Parsing (2008)
Carreras, Xavier, Collins, Michael, Koo, Terry
We describe a parsing approach that makes use of the perceptron algorithm, in conjunction with dynamic programming methods, to recover full constituent-based parse trees. The formalism allows a rich...
Collins, Michael, Globerson, Amir, Koo, Terry, Carreras, Xavier, Bartlett, Peter
Log-linear and maximum-margin models are two commonly-used methods in supervised machine learning, and are frequently used in structured prediction problems. Efficient learning of parameters in these...
A Projected Subgradient Method for Scalable Multi-Task Learning (2008)
Quattoni, Ariadna, Carreras, Xavier, Collins, Michael, Darrell, Trevor
Recent approaches to multi-task learning have investigated the use of a variety of matrix norm regularization schemes for promoting feature sharing across tasks.In essence, these approaches aim at...
A Projected Subgradient Method for Scalable Multi-Task Learning (2008)
Quattoni, Ariadna, Carreras, Xavier, Collins, Michael, Darrell, Trevor
Recent approaches to multi-task learning have investigated the use of a variety of matrix norm regularization schemes for promoting feature sharing across tasks.In essence, these approaches aim at...
Diurnal variation of axial length, intraocular pressure, and anterior eye biometrics (2008)
Read, Scott A., Collins, Michael, Iskander, Daoud Robert
Purpose: To investigate the diurnal variation in axial length and anterior eye biometrics, whilst simultaneously measuring intraocular pressure (IOP) with dynamic contour tonometry in human...
Simple Semi-supervised Dependency Parsing (2008)
Koo, Terry, Carreras, Xavier, Collins, Michael
We present a simple and effective semisupervised method for training dependency parsers. We focus on the problem of lexical representation, introducing features that incorporate word clusters derived...
Joseph G. Davis, Suresh Konda, Helen Granger, Michael Collins, Arthur W. Westerberg
The provision of computer support for collaborative work is a central concern for Information Systems (IS) research and practice. In this paper we present the details of an information flow study...
Michael Collins, Carrie Gates, Gaurav Kataria, Heinz School
We segregate attacks into two categories – targeted and opportunistic – based on whether the attacker compromises a specific target (targeted) or a number of intermediate targets to fulfill his...
(NLP) problems. In many NLP tasks the objects being modeled are strings, trees, graphs or other discrete structures which require some mechanism to convert them into feature vectors. We describe...
Committee Kathleen, R. Mckeown, Michael Collins, Owen Rambow, Advisor Hervé Bourlard, ...
Research Interests Human language technologies and machine learning, in particular, machine translation, automatic summarization, natural language generation, computational models of syntax and...
Measuring Differentiability: Unmasking Pseudonymous Authors (2008)
Jonathan Schler, Elisheva Bonchek-dokow, Michael Collins
In the authorship verification problem, we are given examples of the writing of a single author and are asked to determine if given long texts were or were not written by this author. We present a...
Modeling Sequences Standard tool is the hidden Markov Model (HMM). (2008)
John Lafferty, Doug Beeferman, Adam Berger, Andrew Mccallum, O Pereira, Yoshua Bengio, ...
• Goal: mark up sequences with content tags • Problem: overlapping dependencies on context – long-distance dependencies – multiple levels of granularity (e.g., words & characters) –...
Transfer learning for image classification with sparse prototype representations (2008)
Quattoni, Ariadna, Collins, Michael, Darrell, Trevor
To learn a new visual category from few examples, prior knowledge from unlabeled data as well as previous related categories may be useful. We develop a new method for transfer learning which...
Transfer learning for image classification with sparse prototype representations (2008)
Quattoni, Ariadna, Collins, Michael, Darrell, Trevor
To learn a new visual category from few examples, prior knowledge from unlabeled data as well as previous related categories may be useful. We develop a new method for transfer learning which...
Chris Holme, Leisa Armstrong, Michael Collins
ABSTRACT: There is a need for cost effective tools and data collection methods for field measurements: to increase both productivity and volumes of collected data in the quest for enhanced...
Exponentiated gradient algorithms and PAC-Bayesian (2008)
Peter L. Bartlett, Michael Collins, David Mcallester, Ben Taskar
Large margin methods for structured classification:
Michael Collins, Amir Globerson, Terry Koo, Xavier Carreras, Peter L. Bartlett
Log-linear and maximum-margin models are two commonly-used methods in supervised machine learning, and are frequently used in structured prediction problems. Efficient learning of parameters in these...
Michael Collins, Amir Globerson, Terry Koo, Xavier Carreras, Peter L. Bartlett
Log-linear and maximum-margin models are two commonly-used methods in supervised machine learning, and are frequently used in structured prediction problems. Efficient learning of parameters in these...
A Generalization of Principal ComponentAnalysis to the Exponential Family (2008)
Michael Collins, Sanjoy Dasgupta, Robert E. Schapireat, T Labs
Abstract Principal component analysis (PCA) is a commonly applied techniquefor dimensionality reduction. PCA implicitly minimizes a squared loss function, which may be inappropriate for data that is...
– Parsing ¯ Machine translation Common Themes (2008)
¯ Information extraction (Named entities, Relationships between entities, etc.) ¯ Finding linguistic structure
Steven Abney, Michael Collins, Amit Singhal, Answer Extraction In, Sasha Blair-goldensohn, Kathleen R. Mckeown, ...
gov/projects/duc/roadmapping.html.
(NLP) problems. In many NLP tasks the objects being modeled are strings, trees, graphs or other discrete structures which require some mechanism to convert them into feature vectors. We describe...
fflInformation extraction-Named entities-Relationships between entities fflFinding linguistic structure-P art-of-speech tagging-P arsing fflMachine translation
Learning visual representations using images with captions (2008)
Ariadna Quattoni, Michael Collins, Trevor Darrell
Current methods for learning visual categories work well when a large amount of labeled data is available, but can run into severe difficulties when the number of labeled examples is small. When...
Brian Roark, Murat Saraclar, Michael Collins
This paper describes discriminative language modeling for a large vocabulary speech recognition task. We contrast two parameter estimation methods: the perceptron algorithm, and a method based on...
Diurnal variation of axial length, intraocular pressure, and anterior eye biometrics (2008)
Read, Scott A., Collins, Michael, Iskander, Daoud Robert
Purpose: To investigate the diurnal variation in axial length and anterior eye biometrics, whilst simultaneously measuring intraocular pressure (IOP) with dynamic contour tonometry in human...
Diurnal variation of axial length, intraocular pressure, and anterior eye biometrics (2008)
Read, Scott A., Collins, Michael, Iskander, Daoud Robert
Purpose: To investigate the diurnal variation in axial length and anterior eye biometrics, whilst simultaneously measuring intraocular pressure (IOP) with dynamic contour tonometry in human...
Rabindranath Tagore and Nationalism: An Interpretation (2008)
Rabindranath Tagore is often referred to as a nationalist poet or a nationalist leader. This presents problems both historical and historiographical, since by the end of the first decade of...
Michael Collins, Amir Globerson, Terry Koo, Xavier Carreras, Peter L. Bartlett, John Lafferty
Log-linear and maximum-margin models are two commonly-used methods in supervised machine learning, and are frequently used in structured prediction problems. Efficient learning of parameters in these...
TAG, Dynamic Programming, and the Perceptron for Efficient, Feature-rich Parsing (2008)
Xavier Carreras, Michael Collins, Terry Koo
We describe a parsing approach that makes use of the perceptron algorithm, in conjunction with dynamic programming methods, to recover full constituent-based parse trees. The formalism allows a rich...
Simple semi-supervised dependency parsing (2008)
Terry Koo, Xavier Carreras, Michael Collins
We present a simple and effective semisupervised method for training dependency parsers. We focus on the problem of lexical representation, introducing features that incorporate word clusters derived...
Diurnal variation of axial length, intraocular pressure, and anterior eye biometrics (2008)
Read, Scott A., Collins, Michael, Iskander, Daoud Robert
Purpose: To investigate the diurnal variation in axial length and anterior eye biometrics, whilst simultaneously measuring intraocular pressure (IOP) with dynamic contour tonometry in human...
Michael Collins, Jared Saia, Ling Yu
Aerial search and rescue (or bombing) missions frequently have to locate a target within some prescribed area, running a search pattern that is constrained by the flight characteristics of the...
Theory, 2000. Logistic Regression, AdaBoost and Bregman Distances (2007)
Michael Collins, Robert E. Schapire, Yoram Singer
An extended abstract of this journal submission
Lexicalized PCFG: Parsing Czech (2007)
Michael Collins, Lance Ramshaw, Christoph Tillmann, Jan Hajic
this paper). The sentences in test data are (of course, for a fair test) left unchanged in order. Figure 2: The baseline approach for non-terminal labels. Each label is XP, where X is the POS tag for...
1-57586-150-X, $19.95 Reviewed by (2007)
notes number 88), 1998, xiii+168 pp;
Boosting and the Voted Perceptron (2007)
Under consideration for other conferences (specify)? None. We describe algorithms that rerank the top N hypotheses from a maximum-entropy tagger, the application being named-entity recognition in a...
Abstract A fundamental problem in statistical parsing is the choice of criteria and algorithms used to estimate the parameters in a model. The predominant approach in computational linguistics has...
1 IMPROVING INTONATIONAL PHRASING WITH SYNTACTIC INFORMATION (2007)
Steven Abney, Julia Hirschberg, Michael Collins
The prediction of intonational phrase boundaries from raw text is an important step for a text-to-speech system: Locating where to place short pauses enables more natural sounding speech, that can be...
Signed: ___________________ (2007)
Michael Collins, Michael Collins, Michael Collins
I declare that the work described in this dissertation is, except where otherwise stated, entirely my own work and has not been submitted as an exercise for a degree at this or any other university.
Peter L. Bartlett, Michael Collins, David Mcallester, Ben Taskar
Large margin methods for structured classification:
Michael Collins, Florham Park, Nigel Duffy
This paper introduces new learning algorithms for natural language processing based on the perceptron algorithm. We show how the algorithms can be efficiently applied to exponential sized...
Hodges, Richard E., Kodis, Mary Anne, Epp, Larry W., Orr, Richard, Schuchman, Leonard, Collins, Michael, ...
This paper summarizes recent advances in antenna and power systems technology to enable a high data rate Ka-band Mars-to-Earth telecommunications system. Promising antenna technologies are...
Learning Visual Representations using Images with Captions (2007)
Ariadna Quattoni, Michael Collins, Trevor Darrell
Current methods for learning visual categories work well when a large amount of labeled data is available, but can run into severe difficulties when the number of labeled examples is small. When...
Structured prediction models via the matrix-tree theorem (2007)
Terry Koo, Amir Globerson, Xavier Carreras, Michael Collins
This paper provides an algorithmic framework for learning statistical models involving directed spanning trees, or equivalently non-projective dependency structures. We show how partition functions...
Exponentiated gradient algorithms for log-linear structured prediction (2007)
Amir Globerson, Terry Y. Koo, Xavier Carreras, Michael Collins
Conditional log-linear models are a commonly used method for structured prediction. Efficient learning of parameters in these models is therefore an important problem. This paper describes an...
Online learning of relaxed CCG grammars for parsing to logical form (2007)
Luke S. Zettlemoyer, Michael Collins
We consider the problem of learning to parse sentences to lambda-calculus representations of their underlying semantics and present an algorithm that learns a weighted combinatory categorial grammar...
Online learning of relaxed CCG grammars for parsing to logical form (2007)
Luke S. Zettlemoyer, Michael Collins
We consider the problem of learning to parse sentences to lambda-calculus representations of their underlying semantics and present an algorithm that learns a weighted combinatory categorial grammar...
Structured prediction models via the matrix-tree theorem (2007)
Terry Koo, Amir Globerson, Xavier Carreras, Michael Collins
This paper provides an algorithmic framework for learning statistical models involving directed spanning trees, or equivalently non-projective dependency structures. We show how partition functions...
Morte di uno scrittore, Michael Collins. Traduzione dall'inglese di Seba Pezzani. . - Vicenza. NALUAF000854, Neri Pozza Editore. NAEDAF003782. 2007.
Predicting future botnet addresses with uncleanliness (2007)
Michael Collins, Timothy J. Shimeall, Sidney Faber, Jeff Janies, Rhiannon Weaver, Markus De Shon
The increased use of botnets as an attack tool and the awareness attackers have of blocking lists leads to the question of whether we can effectively predict future bot locations. To that end, we...
Predictors of vinorelbine pharmacokinetics and pharmacodynamics in patients with cancer (2006)
Wong, Mark, Balleine, Rosemary L., Hoskins, Janelle M., Mann, Graham J., Clarke, Christine L., Gurney, Howard, ...
Purpose: Marked interindividual variation in drug disposition and toxicity pose an ongoing challenge to chemotherapy dosage individualization. The aim of this study was to evaluate pretreatment...
Security issues with pervasive computing frameworks (2006)
Michael Collins, Simon Dobson, Paddy Nixon
Abstract. The deployment of pervasive computing systems inevitably raises security concerns. The ability of such systems to synthesise partial information and react to situations without explicit...
Security issues with pervasive computing frameworks (2006)
Michael Collins, Simon Dobson, Paddy Nixon
Abstract. The deployment of pervasive computing systems inevitably raises security concerns. The ability of such systems to synthesise partial information and react to situations without explicit...
A Model for Opportunistic Network Exploits: The Case of P2P Worms (2006)
Michael Collins, Carrie Gates, Gaurav Kataria, Heinz School
We segregate attacks into two categories -- targeted and opportunistic -- based on whether the attacker compromises a specific target (targeted) or a number of intermediate targets to fulfill his end...
Networked Systems Survivability and Assurance (2006)
Public Key Infrastructure, Michael Collins, Technical Analyst, Rae Mcquade, Dede Kirby
for the Department of Energy This document presents the results of an independent assessment by the Sandia National Laboratories Information Design Assurance Red Team of the “NAESB Wholesale...
Predictors of vinorelbine pharmacokinetics and pharmacodynamics in patients with cancer (2006)
Wong, Mark, Balleine, Rosemary L., Hoskins, Janelle M., Mann, Graham J., Clarke, Christine L., Gurney, Howard, ...
Purpose: Marked interindividual variation in drug disposition and toxicity pose an ongoing challenge to chemotherapy dosage individualization. The aim of this study was to evaluate pretreatment...
Komunyakaa's Pictures of Choice: An Introduction (2005)
Callaloo - Volume 28, Number 3, Summer 2005
Komunyakaa, Collaboration, and the Wishbone: An Interview (2005)
Collins, Michael., Komunyakaa, Yusef.
Callaloo - Volume 28, Number 3, Summer 2005
On the Phone with Composer Bill Banfield (2005)
Callaloo - Volume 28, Number 3, Summer 2005
Yusef Komunyakaa: A Bibliography (2005)
Callaloo - Volume 28, Number 3, Summer 2005
What it Means When We Say 'Creole:' an Interview with Salikoko S. Mufwene (2005)
Collins, Michael., Mufwene, Salikoko S.
Callaloo - Volume 28, Number 2, Spring 2005
Clause restructuring for statistical machine translation (2005)
We describe a method for incorporating syntactic information in statistical machine translation systems. The first step of the method is to parse the source language string that is being translated....
Semi-supervised learning for natural language (2005)
Percy Liang, Michael Collins, Percy Liang
Statistical supervised learning techniques have been successful for many natural language processing tasks, but they require labeled datasets, which can be expensive to obtain. On the other hand,...
Object Recognition with Latent Conditional Random Fields (2005)
Ariadna Quattoni, Michael Collins, Ariadna Quattoni
In this thesis we present a discriminative part-based approach for the recognition of object classes from unsegmented cluttered scenes. Objects are modelled as flexible constellations of parts...
Clause restructuring for statistical machine translation (2005)
We describe a method for incorporating syntactic information in statistical machine translation systems. The first step of the method is to parse the source language string that is being translated....
Discriminative syntactic language modeling for speech recognition (2005)
We describe a method for discriminative training of a language model that makes use of syntactic features. We follow a reranking approach, where a baseline recogniser is used to produce 1000-best...
Semi-supervised learning for natural language (2005)
Michael Collins, Percy Liang, Percy Liang
... f--
Hidden-Variable Models for Discriminative Reranking (2005)
We describe a new method for the representation of NLP structures within reranking approaches. We make use of a conditional log–linear model, with hidden variables representing the assignment of...
Discriminative syntactic language modeling for speech recognition (2005)
Michael Collins, Brian Roark, Murat Saraclar
We describe a method for discriminative training of a language model that makes use of syntactic features. We follow a reranking approach, where a baseline recogniser is used to produce 1000-best...
Case-Factor Diagrams for Structured Probabilistic Modeling (2004)
McAllester, David, Collins, Michael, Pereira, Fernando C.N.
We introduce a probabilistic formalism subsuming Markov random fields of bounded tree width and probabilistic context free grammars. Our models are based on a representation of Boolean formulas that...
Corrective language modeling for large vocabulary ASR with the perceptron algorithm (2004)
Brian Roark, Murat Saraclar, Michael Collins
This paper investigates error-corrective language modeling using the perceptron algorithm on word lattices. The resulting model is encoded as a weighted finite-state automaton, and is used by...
Incremental parsing with the perceptron algorithm (2004)
This paper describes an incremental parsing approach where parameters are estimated using a variant of the perceptron algorithm. A beam-search algorithm is used during both training and decoding...
Case-factor diagrams for structured probabilistic modeling (2004)
David Mcallester, Michael Collins, Fernando Pereira
We introduce a probabilistic formalism subsuming Markov random fields of bounded tree width and probabilistic context free grammars. Our models are based on a representation of Boolean formulas that...
Michael Collins, David Mcallester, Ben Taskar
Abstract. We consider the problem of structured classification, where the taskis to predict a label y from an input x, and y has meaningful internal structure.Our framework includes supervised...
A fundamental problem in statistical parsing is the choice of criteria and algorithms used to estimate the parameters in a model. The predominant approach in computational linguistics has been to use...
Abstract A fundamental problem in statistical parsing is the choice of criteria and algorithms used to estimate the parameters in a model. The predominant approach in computational linguistics has...
Case-Factor Diagrams for Structured Probabilistic Modeling (2004)
David Mcallester, Michael Collins, Fernando Pereira
We introduce a probabilistic formalism subsuming Markov random fields of bounded tree width and probabilistic context free grammars. Our models are based on a representation of Boolean formulas that...
An Empirical Analysis of Target-Resident DoS Filters (Extended Abstract) (2004)
Michael Collins, Michael K. Reiter
Michael Collins # Michael K. Reiter + Abstract Numerous techniques have been proposed by which an end-system, subjected to a denial-of-service flood, filters the offending traffic. In this paper, we...
Incremental parsing with the perceptron algorithm (2004)
This paper describes an incremental parsing approach where parameters are estimated using a variant of the perceptron algorithm. A beam-search algorithm is used during both training and decoding...
Conditional random fields for object recognition (2004)
Ariadna Quattoni, Michael Collins, Trevor Darrell
We present a discriminative part-based approach for the recognition of object classes from unsegmented cluttered scenes. Objects are modeled as flexible constellations of parts conditioned on local...
Exponentiated gradient algorithms for large-margin structured classification (2004)
Peter L. Bartlett, Ben Taskar, Michael Collins, David Mcallester
We consider the problem of structured classification, where the task is to predict a label y from an input x, and y has meaningful internal structure. Our framework includes supervised training of...
A fundamental problem in statistical parsing is the choice of criteria and algorithms used to estimate the parameters in a model. The predominant approach in computational linguistics has been to use...
Incremental algorithms for hierarchical classification (2004)
Nicolò Cesa-bianchi, Luca Zaniboni, Michael Collins
We study the problem of classifying data in a given taxonomy when classifications associated with multiple and/or partial paths are allowed. We introduce a new algorithm that incrementally learns a...
Charles Sutton, Andrew Mccallum, Khashayar Rohanimanesh, Michael Collins
In sequence modeling, we often wish to represent complex interaction between labels, such as when performing multiple, cascaded labeling tasks on the same sequence, or when long-range dependencies...
More NetFlow tools: For performance and security (2004)
Carrie Gates, Michael Collins, Michael Duggan, Andrew Kompanek, Mark Thomas
Analysis of network traffic is becoming increasingly important, not just for determining network characteristics and anticipating requirements, but also for security analysis. Several tool sets have...
Anime perse, Michael Collins, traduzione di Massimo Ortelio. . - Vicenza. NALUAF000854, Neri Pozza Editore. NAEDAF003782, 2004.
Anime perse, Michael Collins, traduzione di Massimo Ortelio. . - Vicenza. NALUAF000854, Neri Pozza. NAEDAF003780, 2004.
Anime perse, Michael Collins, traduzione di M. Ortelio. . - Vicenza. NALUAF000854, Neri Pozza. NAEDAF003780, 2004.
Anime perse, Michael Collins, traduzione di Massimo Ortelio. . - Vicenza. NALUAF00001195, Neri Pozza Editore. NAEDAF003782, 2004.
An empirical analysis of target-resident dos filters (2004)
Michael Collins, Michael K. Reiter
Numerous techniques have been proposed by which an end-system, subjected to a denial-of-service flood, filters the offending traffic. In this paper, we provide an empirical analysis of several such...
Head-Driven Statistical Models for Natural Language Parsing (2003)
This article describes three statistical models for natural language parsing. The models extend methods from probabilistic context-free grammars to lexicalized grammars, leading to approaches in...
Discriminative Reranking for Natural Language Parsing (2003)
This paper considers approaches which rerank the output of an existing probabilistic parser. The base parser produces a set of candidate parses for each input sentence, with associated probabilities...
Head-Driven Statistical Models for Natural Language Parsing (2003)
This paper describes three statistical models for natural language parsing. The models extend methods from Probabilistic Context Free Grammars to lexicalized grammars, leading to approaches where a...
A Multi-Lingual Web-Based Survey Form Machine Translation Mechanism (2003)
Richard Boddington, Judy Clayden, Michael Collins, Sam Pride
Translation costs restrict the preparation of medical survey questionnaires in migrant and Aboriginal communities in Western Australia. This is further compounded by a lack of affordable and accurate...
L'altra verità, Michael Collins, traduzione di L. Pugliese. . - Vicenza. NALUAF000854, Neri Pozza. NAEDAF003780, 2003.
I risorti, Michael Collins, traduzione di Massimo Ortelio. . - Vicenza. NALUAF000854, Neri Pozza Editore. NAEDAF003782, 2003.
Soliloquy : a collection of poems / (2002)
Submitted in partial fulfillment of the requirements for the Master of Fine Arts degree.
We describe new algorithms for training tagging models, as an alternative to maximum-entropy models or conditional random elds (CRFs). The algorithms rely on Viterbi decoding of training examples,...
This paper introduces new learning algorithms for natural language processing based on the perceptron algorithm. We show how the algorithms can be efficiently applied to exponential sized...
Ranking Algorithms for Named-Entity Extraction: Boosting and the Voted Perceptron (2002)
This paper describes algorithms which rerank the top N hypotheses from a maximum-entropy tagger, the application being the recovery of named-entity boundaries in a corpus of web data. The first...
Ranking Algorithms for Named-Entity Extraction: Boosting and the Voted Perceptron (2002)
This paper describes algorithms which rerank the top N hypotheses from a maximum-entropy tagger, the application being the recovery of named-entity boundaries in a corpus of web data. The first...
Reranking an N-Gram Supertagger (2002)
John Chen, Srinivas Bangalore, Michael Collins, Owen Rambow
this paper, we investigate an approach to such a choice based on reranking a set of candidate supertags and their confidence scores. RankBoost (Freund et al., 1998) is the boosting algorithm that we...
We describe new algorithms for training tagging models, as an alternative to maximum-entropy models or conditional random fields (CRFs). The algorithms rely on Viterbi decoding of training examples,...
Reranking an n-gram supertagger (2002)
John Chen, Srinivas Bangalore, Michael Collins, Owen Rambow
As shown by Srinivas (1997), standard n-gram modeling may be used to perform supertag disambiguation with accuracy that is adequate for partial parsing, but in general not sufficient for full...
A Generalization of Principal Component Analysis to the Exponential Family (2002)
Michael Collins, Sanjoy Dasgupta, Robert E. Schapire
Principal component analysis (PCA) is a commonly applied technique for dimensionality reduction. PCA implicitly minimizes a squared loss function, which may be inappropriate for data that is not...
A generalization of principal component analysis to the exponential family (2001)
Michael Collins, Sanjoy Dasgupta, Robert E. Schapire
Principal component analysis (PCA) is a commonly applied technique for dimensionality reduction. PCA implicitly minimizes a squared loss function, which may be inappropriate for data that is not...
Convolution Kernels for Natural Language (2001)
We describe the application of kernel methods to Natural Language Processing (NLP) problems. In many NLP tasks the objects being modeled are strings, trees, graphs or other discrete structures which...
A generalization of principal component analysis to the exponential family (2001)
Michael Collins, Sanjoy Dasgupta, Robert E. Schapire
Principal component analysis (PCA) is a commonly applied technique for dimensionality reduction. PCA implicitly minimizes a squared loss function, which may be inappropriate for data that is not...
Convolution Kernels for Natural Language (2001)
We describe the application of kernel methods to Natural Language Processing (NLP) problems. In many NLP tasks the objects being modeled are strings, trees, graphs or other discrete structures which...
A generalization of principal component analysis to the exponential family (2001)
Michael Collins, Sanjoy Dasgupta, Robert E. Schapire
Principal component analysis (PCA) is a commonly applied technique for dimensionality reduction. PCA implicitly minimizes a squared loss function, which may be inappropriate for data that is not...
A generalization of principal component analysis to the exponential family (2001)
Michael Collins, Sanjoy Dasgupta, Robert E. Schapire
Principal component analysis (PCA) is a commonly applied technique for dimensionality reduction. PCA implicitly minimizes a squared loss function, which may be inappropriate for data that is not...
Convolution Kernels for Natural Language (2001)
We describe the application of kernel methods to Natural Language Processing (NLP) problems. In many NLP tasks the objects being modeled are strings, trees, graphs or other discrete structures which...
A generalization of principal component analysis to the exponential family (2001)
Michael Collins, Sanjoy Dasgupta, Robert E. Schapire
Principal component analysis (PCA) is a commonly applied technique for dimensionality reduction. PCA implicitly minimizes a squared loss function, which may be inappropriate for data that is not...
L'altra verità, Michael Collins, traduzione di Luciana Pugliese. . - Vicenza. NALUAF000854, Neri Pozza Editore. NAEDAF003782, 2001.
L'altra verità, Michael Collins, traduzione di Luciana Pugliese. . - Vicenza. NALUAF000854, Neri Pozza Editore. NAEDAF003782, 2001.
LAFITE, Robert, SHIMWELL, Susan, GROCHOWSKI, Nicolas, DUPONT, Jacques, NASH, Linda, SALOMON, Jean-Claude, ...
Suspended Particulate Matter (SPM) concentrations at various levels within the water column, together with salinity and temperature, were measured using water samples collected from six stations...
Improving Open Web Architectures (2000)
When people use the Internet today, they use their browsers to connect to a web server located anywhere in the world and download a specified page that they have requested. Unless this page contains...
Steven Abney, Michael Collins, Amit Singhal
Information retrieval systems have typically concentrated on retrieving a set of documents which are relevant to a user’s query. This paper describes a system that attempts to retrieve a much...
Michael Collins, Robert E. Schapire, Yoram Singer
We give a unified account of boosting and logistic regression in which each learning problem is cast in terms of optimization of Bregman distances. The striking similarity of the two problems in this...
Steven Abney, Michael Collins, Amit Singhal
Information retrieval systems have typically concen-trated on retrieving a set of documents which are rel-evant to a user's query. This paper describes a sys-tem that attempts to retrieve a much...
Discriminative Reranking for Natural Language Parsing (2000)
This paper considers approaches which rerank the output of an existing probabilistic parser. The base parser produces a set of candidate parses for each input sentence, with associated probabilities...
Amit Singhal, Steve Abney, Michiel Bacchiani, Michael Collins, Donald Hindle, O Pereira
In 1999, AT&T participated in the ad-hoc task and the Question Answering (QA), Spoken Document Retrieval (SDR), and Web tracks. Most of our effort for TREC-8 focused on the QA and SDR tracks....
Logistic regression, adaboost and bregman distances (2000)
Michael Collins, Robert E. Schapire, Yoram Singer
Abstract. We give a unified account of boosting and logistic regression in which each learning problem is cast in terms of optimization of Bregman distances. The striking similarity of the two...
The Rules Behind Roles: Identifying Speaker Role in Radio Broadcasts (2000)
Regina Barzilay, Michael Collins, Julia Hirschberg, Steve Whittaker
Previous work has shown that providing information about story structure is critical for browsing audio broadcasts. We investigate the hypothesis that Speaker Role is an important cue to story...
Steven Abney, Michael Collins, Amit Singhal
Information retrieval systems have typically concentrated on retrieving a set of documents which are relevant to a user's query. This paper describes a system that attempts to retrieve a much...
Logistic Regression, AdaBoost and Bregman Distances (2000)
Michael Collins, Robert E. Schapire, Yoram Singer
. We give a unified account of boosting and logistic regression in which each learning problem is cast in terms of optimization of Bregman distances. The striking similarity of the two problems in...
Logistic Regression, AdaBoost and Bregman Distances (2000)
Michael Collins, Robert E. Schapire, Yoram Singer
We give a unified account of boosting and logistic regression in which each learning problem is cast in terms of optimization of Bregman distances. The striking similarity of the two problems in this...
Amit Singhal, Steve Abney, Michiel Bacchiani, Michael Collins, Donald Hindle, O Pereira
In 1999, AT&T participated in the ad-hoc task and the Question Answering (QA), Spoken Document Retrieval (SDR), and Web tracks. Most of our e ort for TREC-8 focused on the QA and SDR tracks....
Logistic Regression, AdaBoost and Bregman Distances (2000)
Michael Collins, Robert E. Schapire, Yoram Singer
We give a unified account of boosting and logistic regression in which each learning problem is cast in terms of optimization of Bregman distances. The striking similarity of the two problems in this...
Capie, Forrest, Collins, Michael
Editada en la Fundación Empresa Pública
Unsupervised Models for Named Entity Classification (1999)
This paper discusses the use of unlabeled examples for the problem of named entity classification. A large number of rules is needed for coverage of the domain, suggesting that a fairly large number...
Unsupervised Models for Named Entity Classification (1999)
mcollins,singer¡ This paper discusses the use of unlabeled examples for the problem of named entity classification. A large number of rules is needed for coverage of the domain, suggesting that a...
Unsupervised Models for Named Entity Classification (1999)
This paper discusses the use of unlabeled examples for the problem of named entity classification. A large number of rules is needed for coverage of the domain, suggesting that a fairly large number...
A Statistical Parser for Czech (1999)
Michael Collins, Lance Ramshaw, Christoph Tillmann
This paper considers statistical parsing of Czech, which differs radically from English in at least two respects: (1) it is a highly inflected language, and (2) it has relatively free word order....
Loevenbruck, Hélène, Collins, Michael, Beckman, Mary, Krishnamurthy, Ashok, Ahalt, Stanley
In the framework of Articulatory Phonology, linguistic structures are represented in terms of coordinated articulatory gestures that are organized into a gestural score. Syllable structure can be...
Loevenbruck, Hélène, Collins, Michael, Beckman, Mary, Krishnamurthy, Ashok, Ahalt, Stanley
In the framework of Articulatory Phonology, linguistic structures are represented in terms of coordinated articulatory gestures that are organized into a gestural score. Syllable structure can be...
Department: English and Comparative Literature.
Risk, Envy and Fear in Sterling Brown's Georgics (1998)
Callaloo - Volume 21, Number 4, Fall 1998
Semantic Tagging using a Probabilistic Context Free Grammar (1998)
This paper describes a statistical model for extraction of events at the sentence level, or "semantic tagging", typi-cally the first level of processing in Information Extrac-tion...
Three Generative, Lexicalised Models for Statistical Parsing (1997)
In this paper we first propose a new statistical parsing model, which is a generative model of lexicalised context-free grammar. We then extend the model to include a probabilistic treatment of both...
Submitted to the Faculty of Science, Technology and Engineering.
Semantic Tagging using a Probabilistic Context Free Grammar (1997)
This paper describes a statistical model for extraction of events at the sentence level, or "semantic tagging", typically the first level of processing in Information Extraction systems. We...
Michael Collins, Jared Saia, Ling Yu
Aerial search and rescue (or bombing) missions frequently have to locate a target within some prescribed area, running a search pattern that is constrained by the flight characteristics of the...
Three Generative, Lexicalised Models for Statistical Parsing (1997)
In this paper we first propose a new statistical parsing model, which is a generative model of lexicalised context-free grammar. We then extend the model to include a probabilistic treatment of both...
this paper gives some background about maximum-likelihood estimation in section 2; considers the major results of DLR, (Wu 83) and (JJ 77) in sections 3, 4 and 5; and concludes in section 6. For a...
A New Statistical Parser Based on Bigram Lexical Dependencies (1996)
This paper describes a new statistical parser which is based on probabilities of dependencies between head-words in the parse tree. Standard bigram probability estimation techniques are extended to...
Prepositional Phrase Attachment through a Backed-Off Model (1995)
Collins, Michael, Brooks, James
Recent work has considered corpus-based or statistical approaches to the problem of prepositional phrase attachment ambiguity. Typically, ambiguous verb phrases of the form {v np1 p np2} are resolved...
Prepositional Phrase Attachment through a Backed-Off Model (1995)
Recent work has considered corpus-based or statistical approaches to the problem of prepositional phrase attachment ambiguity. Typically, ambiguous verb phrases of the form v rip1 p rip2 are resolved...
Prepositional Phrase Attachment through a Backed-Off Model (1995)
Recent work has considered corpus-based or statistical approaches to the problem of prepositional phrase attachment ambiguity. Typically, ambiguous verb phrases of the form v np1 p np2 are resolved...
VDT screen reflections and accommodation response (1994)
Collins, Michael, Davis, Brett, Atchison, David
Reflections in computer screens are imaged behind the screen and therefore potentially create a conflicting cue to appropriate accommodation response to the plane of screen. We investigated the...
The Life and times of a teaboy (1994)
The Life and times of a teaboy, Michael Collins. . - London. NALUAF000421, Phoenix House. NAEDAF004052, 1994.
Calibration of the Canon Autoref R-1 for continuous measurement of accommodation (1993)
Davis, Brett, Collins, Michael, Atchison, David
The Canon Autoref R-1 is an infrared optometer which allows free-space viewing. Pugh and Winn described hardware modifications necessary to convert the Autoref R-1 from an instrument capable of...
Spoken Language Translation With Mid-90's Technology: A Case Study (1993)
Manny Rayner, Ivan Bretan, David Carter, Michael Collins, Vassilios Digalakis, Björn Gambäck, ...
We describe the architecture of the Spoken Language Translator (SLT), a prototype speech translation system which can translate queries from spoken English to spoken Swedish in the domain of air...
Thesis (Ph. D.)--University of Reading, 1990.
Henson, Cynthia A., Collins, Michael, Duke, Stanley H.
Alfalfa (Medicago sativa L. cv. Vernal) nodules were separated into host plant fractions and fractions of rhizobial origin by differential centrifugation and sedimentation equilibrium centrifugation....
El Portador de Fuego: Crónica de un Astronauta / M. Collins; tr. por: Manuel Barberá. (1976)
Traducción de: Carrying the Fire
English commercial banks and business client distress, 1946 63
The article examines the behaviour of two leading English clearing banks, the Midland and Westminster, during periods of financial distress suffered by a sample of business clients during the early...
English industrial distress before 1914 and the response of the banks
Internal bank archives are used to examine the nature of English bank firm relationships during periods of business distress in the decades before 1914. The performance of the banks is examined in...
Effects of Lysozyme and Chitinase on the Spherules of Coccidioides immitis In Vitro
Collins, Michael, Pappagianis, Demosthenes
Spherules of Coccidioides immitis strain Silveira produced in vitro were treated with chitinase and lysozyme. The walls of merthiolate-killed mature endosporulating spherules were degraded by...
Wu, Chia-wei, Glasner, Jeremy, Collins, Michael, Naser, Saleh, Talaat, Adel M.
Infection with Mycobacterium avium subsp. paratuberculosis causes Johne's disease in cattle and is also implicated in cases of Crohn's disease in humans. Another closely related strain, M. avium...
Using Metadata to Generate Web-Based Electronic Data Capture Forms
Collins, Michael, Ross, Eric, Meropol, Neal J., Lazev, Amy B.
Generation of web-based Electronic Data Capture Forms (EDCF) from XML-encoded metadata is an efficient and reliable way to support research data collection. Using metadata-driven development we...
Fleisher, Linda, Millard, Jennifer, Ravitich, Stephanie, Antaramian, Susan, Collins, Michael, Meropol, Neal J.
Badour, Karen, Zhang, Jinyi, Shi, Fabio, Leng, Yan, Collins, Michael, Siminovitch, Katherine A.
Involvement of the Wiskott-Aldrich syndrome protein (WASp) in promoting cell activation requires its release from autoinhibitory structural constraints and has been attributed to WASp association...
The asset portfolio composition of British life insurance firms, 1900 1965
This article examines the investment practices of life assurance firms within the United Kingdom, through an analysis of the asset holdings of the sector over the period 1900 to 1965. The data are...
the article reports the results of primary research aimed at eliciting new information on the nature of the relationship between english banks and the business sector over the long period, 1920 68....
Investor protection and the value effects of bank merger announcements in Europe and the US
Hagendorff, Jens, Collins, Michael, Keasey, Kevin
Investor protection regimes have been shown to partly explain why the same type of corporate event may attract different investor reactions across countries. We compare the value effects of large...
Bank Governance and Acquisition Performance
Jens Hagendorff, Michael Collins, Kevin Keasey
Given the pivotal role of banks in modern economies, the worldwide phenomenon of a high level of bank M&A activity and the consensus view in empirical studies that bank mergers destroy value for...
English industrial distress before 1914 and the response of the banks
Internal bank archives are used to examine the nature of English bank firm relationships during periods of business distress in the decades before 1914. The performance of the banks is examined in...
English commercial banks and business client distress, 1946 63
The article examines the behaviour of two leading English clearing banks, the Midland and Westminster, during periods of financial distress suffered by a sample of business clients during the early...
Bank deregulation and acquisition activity: the cases of the US, Italy and Germany
Jens Hagendorff, Michael Collins, Kevin Keasey
Purpose – Bank regulators across the world have recently lifted restrictions on where banks can operate and what type of activities they can perform. Following the deregulation of the sector, bank...
The Lactobacillus plantarum ftsH Gene Is a Novel Member of the CtsR Stress Response Regulon ▿
Fiocco, Daniela, Collins, Michael, Muscariello, Lidia, Hols, Pascal, Kleerebezem, Michiel, Msadek, Tarek, ...
FtsH proteins have dual chaperone-protease activities and are involved in protein quality control under stress conditions. Although the functional role of FtsH proteins has been clearly established,...