RANLP vol. V

• Volume information (plain text)
• Ordering information
• BibTeX entry for the entire volume [BibTeX entries for all chapters]
• Table of contents [PDF]

Volume Information

↑

Ordering Information

↑

BibTeX Entry

@BOOK{Nicolov-Angelova-Mitkov'2009
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,TITLE     = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}: {S}elected {P}apers from {RANLP} 2007}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Table of Contents

↑

List of Authors and Titles, Abstracts and Individual BibTeX Entries

The paper argues that Guarino is right that ontologies are different from thesauri and similar objects, but not in the ways he believes: they are distinguished from essentially linguistic objects like thesauri and hierarchies of conceptual relations because they unpack, ultimately, in terms of sets of objects and individuals. However this is a lonely status, and without much application outside strict scientific and engineering disciplines, and of no direct relevance to language processing (NLP). More interesting structures, of nlp relevance, that encode conceptual knowledge, cannot be subjected to the "cleaning up" techniques that Guarino advocates, because his conditions are too strict to be applicable, and because the terms used in such structures retain their language-like features of ambiguity and vagueness, and in a way that cannot be eliminated by reference to sets of objects, as it can be in ontologies in the narrow sense. Wordnet is a structure that remains useful to NLP, and has within it features of both types (ontologies and conceptual hierarchies) and its function and usefulness will remain, properly, resistant to Guarino's techniques, because those rest on a misunderstanding about concepts. The ultimate way out of such disputes can only come from automatic construction and evaluation procedures for conceptual and ontological structures from data, which is to say, corpora.

@INCOLLECTION{Wilks'2009
  ,AUTHOR    = {Yorick Wilks}
  ,TITLE     = {{O}ntotherapy, or {H}ow to {S}top {W}orrying {A}bout {W}hat {T}here {I}s}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {1-20}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Goldwasser, Dan, Ming-Wei Chang, Yuancheng Tu & Dan Roth. 2009. "Constraint-Driven Transliteration Discovery". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 21-40. Amsterdam & Philadelphia: John Benjamins.

This paper introduces a novel constraint-driven learning framework for identifying named-entity (ne) transliterations. Traditional approaches to the problem of discovering transliterations depend heav- ily on correctly segmenting the target and the transliteration candidate and on and aligning these segments.
      In this work we propose to formulate the process of aligning segments as a constrained optimization problem. We consider the aligned segments as a latent feature representation and show how to infer an optimal latent representation and how to use it in order to learn an improved discriminative transliteration classifier. Our algorithm is an EM-like iterative algorithm that alternates between an optimization step for the latent representation and a learning step for the classifier's parameters.
      We apply this method both in supervised and unsupervised settings, and show that our model can significantly outperform previous methods trained using considerably more resources.

@INCOLLECTION{Goldwasser-Chang-Tu-Roth'2009
  ,AUTHOR    = {Dan Goldwasser and Ming-Wei Chang and Yuancheng Tu and Dan Roth}
  ,TITLE     = {{C}onstraint-{D}riven {T}ransliteration {D}iscovery}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {21-40}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Menzel, Wolfgang. 2009. "Towards Radically Incremental Parsing of Natural Language". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 41-YY. Amsterdam & Philadelphia: John Benjamins.

Human sentence processing proceeds in a left-to-right incremental manner and is able to assign partial structural interpretations even if all the lexical items involved are not yet known. To mimic such a behaviour in a computational system is particularly difficult for head final languages where the rich information about the valency of the verb comes last. In such cases the Argument Dependency Model facilitates early hypotheses about the thematic status of a constituent by establishing direct argument-to-argument dependencies independent of the verb. An implementation of the model within the frame- work of Weighted Constraint Dependency Grammar (WCDG) is presented and evaluated in a pseudo-incremental processing mode. Due to its non-monotonic nature, the system can replicate the predictions of the Argument Dependency Model about the dynamic nature of revision processes during sentence comprehension with a very high degree of reliability.

@INCOLLECTION{Menzel'2009
  ,AUTHOR    = {Wolfgang Menzel}
  ,TITLE     = {{T}owards {R}adically {I}ncremental {P}arsing of {N}atural {L}anguage}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {41-56}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Sinha, Ravi & Rada Mihalcea. 2009. "Unsupervised Graph-based Word Sense Disambiguation". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 57-72. Amsterdam & Philadelphia: John Benjamins.

This paper describes an unsupervised graph-based method for word sense disambiguation, and presents comparative evaluations using several measures of word semantic similarity and several algorithms for graph centrality. The results indicate that the right combination of similarity metrics and graph centrality algorithms can lead to a performance competing with the state-of-the-art in unsupervised word sense disambiguation, as measured on standard data sets.

@INCOLLECTION{Sinha-Mihalcea'2009
  ,AUTHOR    = {Ravi Sinha and Rada Mihalcea}
  ,TITLE     = {{U}nsupervised {G}raph-based {W}ord {S}ense {D}isambiguation}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {57-72}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Ji, Heng & Ralph Grishman 2009. "Collaborative Entity Extraction and Translation". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 73-84. Amsterdam & Philadelphia: John Benjamins.

Entity extraction is the task of identifying names and nominal phrases (mentions) in a text and linking coreferring mentions. We propose the use of a new source of data for improving entity extraction: the information gleaned from large bitexts and captured by a statistical, phrase-based machine translation system. We translate the individual mentions and test properties of the translated mentions, as well as comparing the translations of coreferring mentions. The results provide feedback to improve source language entity extraction. Experiments on Chinese and English show that this approach can significantly improve Chinese entity extraction (2.2% relative improvement in name tagging F-measure, representing a 15.0% error reduction), as well as Chinese to English entity translation (9.1% relative improvement in F-measure), over state-of-the-art entity extraction and machine translation systems.

@INCOLLECTION{Ji-Grishman'2009
  ,AUTHOR    = {Heng Ji and Ralph Grishman}
  ,TITLE     = {{C}ollaborative {E}ntity {E}xtraction and {T}ranslation}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {73-84}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Patrick Blackburn & Sebastien Hinderer. 2009. "Generating Models for Temporal Representations". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 85-98. Amsterdam & Philadelphia: John Benjamins.

We discuss the use of model building for temporal representations. We chose Polish to illustrate our discussion because it has an interesting aspectual system, but the points we wish to make are not language specific. Rather, our goal is to develop theoretical and computational tools for temporal model building tasks in computational semantics. To this end, we present a first-order theory of time and events which is rich enough to capture interesting semantic distinctions, and an algorithm which takes minimal models for first-order theories and systematically attempts to "perturb" their temporal component to provide non-minimal, but semantically significant, models.

@INCOLLECTION{Blackburn-Hinderer'2009
  ,AUTHOR    = {Patrick Blackburn and Sebastien Hinderer}
  ,TITLE     = {{G}enerating {M}odels for {T}emporal {R}epresentations}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {85-98}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Ramsay, Allan. 2009. "The Complexity of Everyday Language". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 99-112. Amsterdam & Philadelphia: John Benjamins.

Everyday language is remarkably complex. We say very subtle things, and our hearers understand the subtleties and respond appropriately. In order to develop computational systems which can cope with this complexity, we have to use meaning representations that preserve the fine distinctions that we make when we use natural language, and we have to have inference engines that can draw appropriate conclusions from these meaning representations. The current paper provides a reminder of some of these distinctions. See (Ramsay 2001) for the development of the required inference engine.

@INCOLLECTION{Ramsay'2009
  ,AUTHOR    = {Allan Ramsay}
  ,TITLE     = {{T}he {C}omplexity of {E}veryday {L}anguage}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {99-112}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Knights, Dan, Mike Mozer & Nicolas Nicolov. 2009. "Detecting topic drift". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 113-130. Amsterdam & Philadelphia: John Benjamins.

The Latent Dirichlet Allocation (lda) topic model of Blei, Ng, & Jordan (2003) is well-established as an effective approach to recovering meaningful topics of conversation from a set of documents. However, a useful analysis of user-generated content is concerned not only with the recovery of topics from a static data set, but with the evolution of topics over time. In this chapter we discuss the usage of compound topic models (ctms) to track topics across two distinct data sets (i.e., past and present) and to visualize trends in topics over time; we describe several metrics for detecting a change in the distribution of topics within a time-window; and we illustrate how these approaches discover emerging conversation topics related to current events in real data sets. We also give a step-by-step explanation of the Gibbs sampling implementation of lda proposed by Griffiths and Steyvers (2004, 2006).

@INCOLLECTION{Knights-Mozer-Nicolov'2009
  ,AUTHOR    = {Dan Knights and Mike Mozer and Nicolas Nicolov}
  ,TITLE     = {{D}etecting {T}opic {D}rift}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {113-130}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Morante, Roser & Antal van den Bosch. 2009. "Feature Construction for Memory-Based Semantic Role Labeling of Catalan and Spanish". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 131-142. Amsterdam & Philadelphia: John Benjamins.

To improve the performance of a single-classifier memory-based semantic role labeling (SRL) system for Catalan and Spanish, we con- struct new predictive features based on originally multi-valued features. We split and binarize these features, and construct new fea- tures by combining the most informative multi-valued features. The new system is tested on in-domain and out-of-domain corpora, achieving state-of-the-art performance, and error reductions ranging from 6.93 to 20.59 over the original system. The improvements are due to new features constructed out of the two most informative original features, viz. the syntactic function and the preposition of the sibling phrase in focus.

@INCOLLECTION{Morante-van-den-Bosch'2009
  ,AUTHOR    = {Roser Morante and Antal {van den Bosch}}
  ,TITLE     = {{F}eature {C}onstruction for {M}emory-{B}ased {S}emantic {R}ole {L}abeling of {C}atalan and {S}panish}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {131-142}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Boudin, Florian & Juan Manuel Torres-Moreno. 2009. "A Maximization-Minimization Approach for Update Text Summarization". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 143-154. Amsterdam & Philadelphia: John Benjamins.

The work presents an update summarization system that uses a combination of two techniques to generate extractive summaries which focus on new but relevant information. A fast maximization-minimization approach is used to select sentences that are distant from sen- tences used in already read documents and at the same time close to the topic. On top of this sentence scoring approach, a second method called "Novelty Boosting" is used. The latter extends the topic by the unique terms in the update document cluster, thus biasing the cosine maximization-minimization towards maximizing relevance of a summary sentence not only with respect to the topic, but also to the novel aspects of the topic in the update cluster. Results are based on the duc 2007 update summarization task.

@INCOLLECTION{Boudin-Torres-Moreno'2009
  ,AUTHOR    = {Florian Boudin and Juan Manuel Torres-Moreno}
  ,TITLE     = {{A} {M}aximization-{M}inimization {A}pproach for {U}pdate {T}ext {S}ummarization}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {143-154}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Çetinoğlu, Özlem & Kemal Oflazer. 2009. "Integrating Derivational Morphology into Syntax". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 155-YY. Amsterdam & Philadelphia: John Benjamins.

In this chapter we discuss the use of intermediate lexical units - inflectional groups - to handle the productive derivational morphology, in the development of a lexical functional grammar for Turkish. With the help of such parsing units, it is possible to reflect the consequences of the derivations to c- and f-structure representations with succinct and manageable rules. We first explain how we use these parsing units for handling simple productive derivational morphology and then focus on linguistically more interesting cases such as long distance dependencies and complex predicates. Free constituent order and combining the morphemes carrying different pieces of semantic information into one semantic form, make the implementation computationally more challenging. We mainly investigate how our approach affects handling several linguistic phenomena.

@INCOLLECTION{Cetinoglu-Oflazer'2009
  ,AUTHOR    = {\"{O}zlem \c{C}et\.{i}no\u{g}lu and Kemal Oflazer}
  ,TITLE     = {{I}ntegrating {D}erivational {M}orphology into {S}yntax}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {155-170}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Jiampojamarn, Sittichai, Grzegorz Kondrak & Colin Cherry. 2009. "Biomedical Named Entity Recognition Using Discriminative Training". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 171-180. Amsterdam & Philadelphia: John Benjamins.

We demonstrate the efficiency of the online discriminative training framework by applying it to the biomedical named entity recognition problem. We show that discriminative training overcomes the limited expressiveness of the traditional, generative hmms by incorporating additional, potentially overlapping features. This simple and elegant learning method produces performance that is comparable to the current state-of-the-art, while using only straightforward features derived from the provided training data. We investigate two online update techniques: averaged perceptron and Margin Infused Relaxed Algorithm (MIRA). Our experiments illustrate the relative value of competing techniques that employ more complex learning algorithms and semantic features constructed from external resources.

@INCOLLECTION{Jiampojamarn-Kondrak-Cherry'2009
  ,AUTHOR    = {Sittichai Jiampojamarn and Grzegorz Kondrak and Colin Cherry}
  ,TITLE     = {}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {171-180}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Adafre, Sisay Fissaha, Maarten de Rijke & Erik Tjong Kim Sang. 2009. "Completing Lists of Entities". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 181-192. Amsterdam & Philadelphia: John Benjamins.

We consider the list completion task, an entity retrieval task where, in return to a topic statement and a number of example entities, systems have to return further examples. For this task, we propose and evaluate several algorithms. One of the core challenges is to overcome the very limited amount of information that serves as input - to address this challenge we explore different representations of list descriptions. For evaluation purposes we make use of the lists and categories available in Wikipedia. Experimental results show that cluster-based contexts improve retrieval results.

@INCOLLECTION{Adafre-de-Rijke-Sang'2009
  ,AUTHOR    = {Sisay Fissaha Adafre and Maarten de Rijke and Erik Tjong Kim Sang}
  ,TITLE     = {{C}ompleting {L}ists of {E}ntities}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {181-192}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Vilares, Jesús, Michael P. Oakes & Manuel Vilares. 2009. "Character N-Grams as Text Alignment Unit: CLIR Applications". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 193-204. Amsterdam & Philadelphia: John Benjamins.

This article presents a new knowledge-light approach for query translation in Cross-Language Information Retrieval. Our approach is in- spired by previous similar work, preserving its advantages but avoiding its main drawbacks. It is based on the direct translation of character n-grams through subword-level text alignment, thus avoiding the need for word normalization, and also dealing with out-of- vocabulary words. Moreover, since such a solution does not rely on language-specific processing, it can be used when linguistic information and resources are scarce or unavailable. Our system has been tested using the CLEF evaluation corpus.

@INCOLLECTION{Vilares-Oakes-Vilares'2009
  ,AUTHOR    = {Jes\'{u}s Vilares and Michael P. Oakes and Manuel Vilares}
  ,TITLE     = {{C}haracter {N}-{G}rams as {T}ext {A}lignment {U}nit: {CLIR} {A}pplications}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {193-204}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Choi, Jinho D. & Nicolas Nicolov. 2009. "K-best, locally pruned, transition-based dependency parsing using robust risk minimization". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 205-216. Amsterdam & Philadelphia: John Benjamins.

We combine transition-based dependency parsing with a high- performing but relatively underexplored machine learning technique, Robust Risk Minimization. During decoding, we judiciously prune the next parsing states using k-best ranking. Moreover, we apply a simple post-processing to ensure robustness. We evaluate our approach on the CoNLL'09 shared task English data and improve transition-based dependency parsing accuracy, reaching a labeled attachment score of 89.28%. We also have observed near quadratic average running time in practice for the algorithm.

@INCOLLECTION{Choi-Nicolov'2009
  ,AUTHOR    = {Jinho D. Choi and Nicolas Nicolov}
  ,TITLE     = {{K}-best, {L}ocally {P}runed, {T}ransition-based {D}ependency {P}arsing {U}sing {R}obust {R}isk {M}inimization}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {205-216}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Field, Debora & Allan Ramsay. 2009. "Minimal Sets of Minimal Speech Acts". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 217-226. Amsterdam & Philadelphia: John Benjamins.

Work on speech acts has generally involved the introduction of sets of different actions such as informing, reminding, bluffing, and lying, which have different preconditions and effects, and hence can be used to achieve a wide variety of different real-world goals. They tend to have indistinguishable surface forms, however. As such, it is extremely hard for the hearer to decide which action she thinks has been performed. It is therefore also extremely difficult for the speaker to be confident about how the hearer will respond. We will show how to achieve complex goals on the basis of a very simple set of linguistic actions. These actions have clearly marked surface forms, and hence can easily be distinguished by a hearer. In order to do this, we have developed an epistemic planner with several interesting features, and with several optimisations that relate directly to aspects of the task at hand.

@INCOLLECTION{Field-Ramsay'2009
  ,AUTHOR    = {Debora Field and Allan Ramsay}
  ,TITLE     = {Minimal Sets of Minimal Speech Acts}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {217-226}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Islam, Aminul & Diana Inkpen. 2009. "Semantic Similarity of Short Texts". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 227-236. Amsterdam & Philadelphia: John Benjamins.

This chapter presents a method for measuring the semantic similarity of texts using a corpus based measure of semantic word similarity and a normalized and modified versions of the Longest Common Subsequence (LCS) string matching algorithm. Existing methods for computing text similarity have focused mainly on similarity of documents or similarity of words. In this paper, we focus on computing the similarity between two sentence or between two short paragraphs. The proposed method can be exploited in a variety of applications involving textual knowledge representation and knowledge discovery.

@INCOLLECTION{Islam-Inkpen'2009
  ,AUTHOR    = {Aminul Islam and Diana Inkpen}
  ,TITLE     = {{S}emantic {S}imilarity of {S}hort {T}exts}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {227-236}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Tiedemann, Jörg. 2009. "News from OPUS - A Collection of Multilingual Parallel Corpora with Tools and Interfaces". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 227-248. Amsterdam & Philadelphia: John Benjamins.

The OPUS corpus is a growing resource providing various multilingual parallel corpora from different domains. In this article we introduce resources that have recently been added to opus. We also look at some corpus-specific problems and the solutions used in preparing the parallel data for the inclusion in our collection. In particular, we discuss the alignment of movie subtitles and the conversion of biomedical documents and localization data to a sentence aligned XML format. opus also includes various tools and interfaces besides the actual data. We will briefly describe our corpus processing and query tools and a newly added lexical database of word alignments.

@INCOLLECTION{Tiedemann'2009
  ,AUTHOR    = {J\"{o}rg Tiedemann}
  ,TITLE     = {{N}ews from {OPUS} -- {A} {C}ollection of {M}ultilingual {P}arallel {C}orpora with {T}ools and {I}nterfaces}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {227-248}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Navarretta, Costanza. 2009. "Reusing Contemporary Language Resources to PoS Tag Non-contemporary Literary Texts". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 249-258. Amsterdam & Philadelphia: John Benjamins.

This paper presents a refinement of the methodology to semi-automatically construct training data for PoS tagging older Danish literary texts belonging to different periods of time presented by Navarretta (2007). The methodology reuses contemporary general language resources. We describe how the methodology has been applied to train the TreeTagger (Schmid 1994) to tag fairy tales from the 19th century and how the tagger has been tested on literary texts written the last two hundred years. The best performance obtained in these experiments is significantly better than that achieved on contemporary texts by the same tagger trained on contemporary data. Our tests show also that the performance of the tagger varies significantly from text type to text type. The methodology can also be used to tune the tagger on particular text types and domains.

@INCOLLECTION{Navarretta'2009
  ,AUTHOR    = {Costanza Navarretta}
  ,TITLE     = {{R}eusing {C}ontemporary {L}anguage {R}esources to {POS} {Tag} {N}on-contemporary {L}iterary {T}exts}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {249-258}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Agerri, Rodrigo, John Barnden, Mark Lee & Alan Wallington. 2009. "Domain Independent Mappings and Source Domain Reasoning for the Semantics of Metaphor". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 259-268. Amsterdam & Philadelphia: John Benjamins.

We present a computationally-oriented formal semantic framework to address the interpretation of metaphorical text. The framework builds on two main features: inference on source domain terms and a set of special mappings. Such mappings are domain-independent and are adjuncts to any conceptual metaphor.

@INCOLLECTION{Agerri-Barnden-Lee-Wallington'2009
  ,AUTHOR    = {Rodrigo Agerri and John Barnden and Mark Lee and Alan Wallington}
  ,TITLE     = {{D}omain {I}ndependent {M}appings and {S}ource {D}omain {R}easoning for the {S}emantics of {M}etaphor}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {259-268}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Catherine Havasi, Robert Speer & Jason Alonso. 2009. "ConceptNet: A Lexical Resource for Common Sense Knowledge". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 269-280. Amsterdam & Philadelphia: John Benjamins.

The Open Mind Common Sense project has been collecting common- sense knowledge from volunteers on the Internet since 2000. This knowledge is represented in a machine-interpretable semantic network called ConceptNet. We present ConceptNet 3, which improves the acquisition of new knowledge in ConceptNet and facilitates turning edges of the network back into natural language. We show how its modular design helps it adapt to different data sets and languages. Finally, we evaluate the content of ConceptNet 3 and show that it contains significant overlaps with WordNet and the Brandeis Semantic Ontology. Using principal component analysis on the graph structure of ConceptNet yields AnalogySpace, a vector space representation of common sense knowledge. This representation reveals large-scale patterns in the data, while smoothing over noise, and predicts new knowledge that the database should contain.

@INCOLLECTION{Havasi-Speer-Alonso'2009
  ,AUTHOR    = {Catherine Havasi and Robert Speer and Jason Alonso}
  ,TITLE     = {{C}oncept{N}et: {A} {L}exical {R}esource for {C}ommon {S}ense {K}nowledge}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {269-280}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Chen, John, Laurie Crist, Len Eynon, Cassandre Creswell, Amit Mhatre & Rohini Srihari. 2009. "Confidence Measures and Thresholding in Coreference Resolution". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 281-290. Amsterdam & Philadelphia: John Benjamins.

Coreference resolution is an important component of information extraction systems. Pairwise coreference systems utilizing machine learning methods have been found to perform quite well for this task. We improve the performance of this kind of system by using a confidence measure in order to filter out low-scoring classifications through thresholding. We find that a multiple threshold system, with the thresholds defined the right way, outperforms both the baseline and a single threshold system. We also find that basing a threshold as close as possible to the evaluation metric is a good idea. We define a novel evaluation metric whose chief attraction is that it does not require hand-annotation of test data. We see that our approach scores well on the new metric.

@INCOLLECTION{Chen-Crist-Eynon-Creswell-Mhatre-Srihari'2009
  ,AUTHOR    = {John Chen and Laurie Crist and Len Eynon and Cassandre Creswell and Amit Mhatre and Rohini Srihari}
  ,TITLE     = {{C}onfidence {M}easures and {T}hresholding in {C}oreference {R}esolution}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {281-290}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Orăsan, Constantin. 2009. "The Influence of Pronominal Anaphora Resolution on Term-based Summarisation". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 291-300. Amsterdam & Philadelphia: John Benjamins.

The assumption of term-based summarisation is that the importance of a sentence can be determined by the importance of the words it contains. One drawback to this approach is that it usually considers words in isolation, ignoring relations between them such as anaphoric links. This paper investigates to what extent the integration of pronominal anaphora resolvers into a term-based summariser can improve the informativeness of the summaries it produces. Evaluation shows that anaphora resolution can have a beneficial effect, but that highly accurate methods are necessary.

@INCOLLECTION{Orasan'2009
  ,AUTHOR    = {Or\u{a}san, Constantin}
  ,TITLE     = {{T}he {I}nfluence of {P}ronominal {A}naphora {R}esolution on {T}erm-based {S}ummarisation}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {291-300}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Tachbelie, Martha Yifiru & Wolfgang Menzel. 2009. "Morpheme-based Language Modeling for an Inflectional Language - Amharic". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 301-310. Amsterdam & Philadelphia: John Benjamins.

This paper presents morpheme-based language models for Amharic, a morphologically rich and under-resourced language. The language models have been developed with different n-gram orders and smoothing techniques. Among the developed models, the best performing ones are the 5-gram and 6-gram models with modified Kneser-Ney smoothing and interpolation of n-gram probability estimates using automatically and manually segmented data, respectively. The morpheme-based models have low perplexity and compared to word-based models, they have a small number of out-of-vocabulary words. It has also been observed that the quality of models depends upon the segmentation quality.

@INCOLLECTION{Tachbelie-Menzel'2009
  ,AUTHOR    = {Martha Yifiru Tachbelie and Wolfgang Menzel}
  ,TITLE     = {{M}orpheme-based {L}anguage {M}odeling for an {I}nflectional {L}anguage -- {A}mharic}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {301-310}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Demir, Seniz, Sandra Carberry & Stephanie Elzer. 2009. "Issues in Realizing the Overall Message of a Bar Chart". Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 311-320. Amsterdam & Philadelphia: John Benjamins.

Information graphics, such as bar charts and line graphs, that appear in popular media generally have a message that they are intended to convey. We have developed a Bayesian network that recognizes the overall message of a bar chart and produces the logical representation of that message. However, the realization of a logical representation in natural language requires constructing referents for certain graphical elements. This chapter presents our solution to one aspect of this problem: identifying an appropriate referent for the dependent axis. An evaluation study validates our methodology and shows that it is much better than several baseline strategies.

@INCOLLECTION{Demir-Carberry-Elzer'2009
  ,AUTHOR    = {Seniz Demir and Sandra Carberry and Stephanie Elzer}
  ,TITLE     = {{I}ssues in {R}ealizing the {O}verall {M}essage of a {B}ar {C}hart}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {311-320}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

Chanev, Atanas, Kiril Simov, Petya Osenova & Svetoslav Marinov. 2009. The BulTreeBank: Parsing and Conversion. Recent Advances in Natural Language Processing V (= Current Issues in Libguistic Theory, CILT 309), pp. 321-330. Amsterdam & Philadelphia: John Benjamins.

Treebanks are often based on either of two grammatical formalisms: phrase structure (constituency) grammar or dependency grammar. However, sometimes it is necessary to transform treebank representations in order to test statistical parsers based on the alternative approach. In this paper we present new parsing results for Bulgarian by training two statistical parsers (constituency and dependency) on the BulTreeBank. We explore the interaction between constituency and dependency representations in both the constituency and the dependency parser using information based on the alternative formal- ism. We show that this interaction has a positive impact on parsing accuracy. We also investigate the relation between the BulTreeBank and one of its dependency variants which had been automatically derived from the original treebank.

@INCOLLECTION{Chanev-Simov-Osenova-Marinov'2009
  ,AUTHOR    = {Atanas Chanev and Kiril Simov and Petya Osenova and Svetoslav Marinov}
  ,TITLE     = {{T}he {BulTreeBank}: {P}arsing and {C}onversion}
  ,BOOKTITLE = {{R}ecent {A}dvances in {N}atural {L}anguage {P}rocessing {V}}
  ,PUBLISHER = {John Benjamins}
  ,SERIES    = {{C}urrent {I}ssues in {L}inguistic {T}heory}
  ,VOLUME    = {309}
  ,EDITOR    = {Nicolas Nicolov and Galia Angelova and Ruslan Mitkov}
  ,PAGES     = {321-330}
  ,YEAR      = {2009}
  ,ADDRESS   = {Amsterdam \& Philadelphia}
}

↑