BibSonomy publications for /concept/user/butonic/nlpMon Jun 04 16:10:49 CEST 2007Proceedings of the European Semantic Web Conference, ESWC2007JulyLecture Notes in Computer Science{Empowering Software Maintainers with Semantic Web Technologies}45192007papers NT2OD future work Mon May 07 12:13:13 CEST 2007Cape Town, South Africa8th World Conference on Computers in EducationTeaching Modeling with Objects First2005se NT2OD fujaba Sat May 05 18:20:19 CEST 2007Proceedings of the Third ACL Workshop on Very Large CorporaText Chunking Using Transformation-Based Learning1995plugin NT2OD gate Sat May 05 18:19:47 CEST 2007Hong KongProceedings of the 38 th Annual Meeting of the Association for Computational LinguisticsOctoberIndependence and commitment: Assumptions for rapid training and execution of rule-based POS taggers2000plugin NT2OD gate Sat May 05 17:21:17 CEST 2007Proceedings of the 40th Anniversary Meeting of the Association for Computational Linguistics{GATE: A framework and graphical development environment for robust NLP tools and applications}2002gate NT2OD Tue May 01 18:55:53 CEST 2007Annual Meeting of the Association for Computational Linguistics423-430Accurate Unlexicalized Parsing412003NT2OD stanford nlp parser We demonstrate that an unlexicalized PCFG can parse much more accurately than previously shown, by making use of simple, linguistically motivated state splits, which break down false independence assumptions latent in a vanilla treebank grammar. Indeed, its performance of 86.36% (LP/LR F1 ) is better than that of early lexicalized PCFG models, and surprisingly close to the current state-of-the-art. This result has potential uses beyond establishing a strong lower bound on the maximum possible accuracy of unlexicalized models: an unlexicalized PCFG is much more compact, easier to replicate, and easier to interpret than more complex lexical models, and the parsing algorithms are simpler, more widely understood, of lower asymptotic complexity, and easier to optimize.Tue May 01 18:52:34 CEST 2007Advances in Neural Information Processing SystemsFast Exact Inference with a Factored Model for Natural Language Parsing152003nlp stanford parser NT2OD We present a novel generative model for natural language tree structures in which semantic (lexical dependency) and syntactic (PCFG) structures are scored with separate models. This factorization provides conceptual simplicity, straightforward opportunities for separately improving the component models, and a level of performance comparable to similar, non-factored models. Most importantly, unlike other modern parsing models, the factored model admits an extremely effective A* parsing algorithm, which enables efficient, exact inference.Tue May 01 16:26:59 CEST 2007Maximum Entropy Models for Natural Language Ambiguity Resolution1998dissertation opennlp maxent NT2OD nlp Sun Mar 25 17:33:33 CEST 2007Tzigov Chark, BulgariaProceedings of the Recent Advances in Natural Language Processing 2001 Conference257-274{N}amed {E}ntity {R}ecognition from {D}iverse {T}ext {T}ypes2001NT2OD 2001 nlp gate Current research in Information Extraction tends to be focused on application-specific systems tailored to a particular domain. The Muse system is a multi-purpose Named Entity recognition system which aims to reduce the need for costly and time-consuming adaptation of systems to new applications, with its capability for processing texts from widely differing domains and genres. Although the system is still under development, preliminary results are encouraging, showing little degradation when processing texts of lower quality or of unusual types. The system currently averages 93% precision and 95% recall across a variety of text types.Sun Mar 25 17:32:18 CEST 2007Proceedings of the IEEE / ACL 2006 Workshop on Spoken Language TechnologyThe ({N}on){U}tility of {L}inguistic {F}eatures for {P}redicting {P}rominence in {S}pontaneous {S}peech2006stanford nlp parsetree parser NT2OD 2006 Conversational speech is characterized by prosodic variability which makes pitch accent prediction for this genre especially difficult. The linguistic literature points out that complex features such as information status, contrast and animacy help predict pitch accent placement. In this paper, we use a corpus annotated for such features to determine if they improve prominence prediction over traditional shallow features such as frequency and part-of-speech, or over new ones that we introduce. We demonstrate that while correlated with prominence, complex linguistic features do not improve prediction accuracy. Furthermore, the performance of our classifier is quite close to the ceiling defined by variability in human accent placement. An oracle experiment demonstrates, though, that at least some accuracy improvement is still possible.Sun Mar 25 17:30:32 CEST 2007Proceedings of the IEEE / ACL 2006 Workshop on Spoken Language Technology{G}enerating {T}yped {D}ependency {P}arses from {P}hrase {S}tructure {P}arses2006parser 2006 nlp parsetree NT2OD stanford This paper describes a system for extracting typed dependency parses of English sentences from phrase structure parses. In order to capture inherent relations occurring in corpus texts that can be critical in real-world applications, many NP relations are included in the set of grammatical relations used. We provide a comparison of our system with Minipar and the Link parser. The typed dependency extraction facility described here is integrated in the Stanford Parser, available for download.Sun Mar 25 17:22:04 CEST 2007Cambridge, MassachusettsFoundations of Statistical Natural Language Processing1999NT2OD nlp Sun Feb 25 20:06:58 CET 2007Dot. Kom Deliverable D 3-3Sep{IE} {E}valuation {S}trategy 2003NT2OD nlp AIFB 2003 This deliverable defines the evaluation strategies for the Information Extraction tools that are under development within the Dot.Kom project.Sun Feb 25 20:01:00 CET 2007Jan{U}sing {GATE} as an {A}nnotation {T}ool2005nlp NT2OD gate 2005 28This manual is designed as an introduction to GATE 3 for people who have no experience at all with the tool. The first part covers the basic aspects of how to use GATE as an annotation tool; the second part includes some more advanced aspects concerned with using the ontology functionalities. For more detailed information, we refer the reader to the GATE User Guide (see Section 3.2).Sun Feb 25 19:59:56 CET 2007Sheffield, UKProceedings of the SIGIR 04 Workshop: Information Retrieval for Question AnsweringJul{D}omain-{S}pecific {QA} for the {C}onstruction {S}ector 2004NT2OD 2004 nlp gate domain 29Previous research on Question-Answering (QA) has focused on general domain questions. The general approach is to first recognize Named Entities (NE) in both texts and questions; then the most relevant answers (or passages) are selected (after an IR selection) according to the type of question and the NEs included in the possible answers. In this paper, we extend this general approach to domain-specific questions in the construction sector. This extension allows us to answer questions such as “What material is appropriate for ...”, which cannot be answered by a general QA system. Our approach is based on a domain-specific thesaurus, which contains a large set of domain-specific concepts organized into a hierarchy. Generic concepts such as “material” are considered as semantic categories. Our experiments on a technical corpus in construction show that this approach is effective: using our extension, we can obtain improvements on Mean Reciprocal Rank (MRR) of about 10%.Sun Feb 25 19:57:21 CET 2007Proceedings of the North American Association of Computational LinguisticsLearning to recognize features of valid textual entailments2006parsetree nlp 2006 NT2OD parser stanford This paper advocates a new architecture for textual inference in which finding a good alignment is separated from evaluating entailment. Current approaches to semantic inference in question answering and textual entailment have approximated the entailment problem as that of computing the best alignment of the hypothesis to the text, using a locally decomposable matching score. We argue that there are significant weaknesses in this approach, including flawed assumptions of monotonicity and locality. Instead we propose a pipelined approach where alignment is followed by a classification step, in which we extract features representing high-level characteristics of the entailment problem, and pass the resulting feature vector to a statis- be seen in this light.) In this paper, we highlight the tical classifier trained on development data. We report results on data from the 2005 Pascal RTE Challenge which surpass previously reported results for alignment-based systems.Sun Feb 25 19:56:57 CET 2007Proceedings of the Conference on Empirical Methods in Natural Language Processing{U}nsupervised {D}iscovery of a {S}tatistical {V}erb {L}exicon2006parser stanford nlp 2006 parsetree NT2OD This paper demonstrates how unsupervised techniques can be used to learn models of deep linguistic structure. Determining the semantic roles of a verb’s dependents is an important step in natural language understanding. We present a method for learning models of verb argument patterns directly from unannotated text. The learned models are similar to existing verb lexicons such as VerbNet and PropBank, but additionally include statistics about the linkings used by each verb. The method is based on a structured probabilistic model of the domain, and unsupervised learning is performed with the EM algorithm. The learned models can also be used discriminatively as semantic role labelers, and when evaluated relative to the PropBank annotation, the best learned model reduces 28% of the error between an informed baseline and an oracle upper bound.Sun Feb 25 19:48:04 CET 2007{D}eveloping {L}anguage {P}rocessing {C}omponents with {GATE} {V}ersion 3 (a {U}ser {G}uide)20052005 NT2OD gate Sun Feb 25 19:34:27 CET 2007Conference on Empirical Methods in Natural Language Processing{S}olving the {P}roblem of {C}ascading {E}rrors: {A}pproximate {B}ayesian {I}nference for {L}inguistic {A}nnotation {P}ipelines2006parsetree stanford 2006 parser NT2OD nlp The end-to-end performance of natural language processing systems for compound tasks, such as question answering and textual entailment, is often hampered by use of a greedy 1-best pipeline archi tecture, which causes errors to propagate and compound at each stage. We present a novel architecture, which models these pipelines as Bayesian networks, with each low level task corresponding to a variable in the network, and then we perform approximate inference to find the best labeling. Our approach is extremely simple to apply but gains the benefits of sampling the entire distribution over labels at each stage in the pipeline. We apply our method to two tasks – semantic role labeling and recognizing textual entailment – and achieve useful performance gains from the superior pipeline architecture. Sun Feb 25 19:33:30 CET 2007Proceedings of the 44 th Annual Meeting of the Association for Computational LinguisticsReceived Best Paper Award{S}emantic {T}axonomy {I}nduction from {H}eterogenous {E}vidence2006best 2006 NT2OD parser stanford parsetree nlp We propose a novel algorithm for inducing semantic taxonomies. Previous algorithms for taxonomy induction have typically focused on independent classifiers for discovering new single relationships based on hand-constructed or automatically discovered textual patterns. By contrast, our algorithm flexibly incorporates evidence from multiple classifiers over heterogenous relationships to optimize the entire structure of the taxonomy, using knowledge of a word’s coordinate terms to help in determining its hypernyms, and vice versa. We apply our algorithm on the problem of sense-disambiguated noun hyponym acquisition, where we combine the predictions of hypernym and coordinate term classifiers with the knowledge in a preexisting semantic taxonomy (WordNet 2.1). We add 10, 000 novel synsets to WordNet 2.1 at 84% precision, a relative error reduction of 70% over a non-joint algorithm using the same component classifiers. Finally, we show that a taxonomy built using our algorithm shows a 23% relative F-score improvementover WordNet 2.1 on an independent testset of hypernym pairs.