Datasets:

MaartenGr

/

arxiv_nlp

like 2

Assessing System Agreement and Instance Difficulty in the Lexical Sample Tasks of Senseval-2

This paper presents a comparative evaluation among the systems that participated in the Spanish and English lexical sample tasks of Senseval-2. The focus is on pairwise comparisons among systems to assess the degree to which they agree, and on measuring the difficulty of the test instances included in these tasks.

Computation and Language

Machine Learning with Lexical Features: The Duluth Approach to Senseval-2

This paper describes the sixteen Duluth entries in the Senseval-2 comparative exercise among word sense disambiguation systems. There were eight pairs of Duluth systems entered in the Spanish and English lexical sample tasks. These are all based on standard machine learning algorithms that induce classifiers from sense-tagged training text where the context in which ambiguous words occur are represented by simple lexical features. These are highly portable, robust methods that can serve as a foundation for more tailored approaches.

Computation and Language

Thumbs up? Sentiment Classification using Machine Learning Techniques

We consider the problem of classifying documents not by topic, but by overall sentiment, e.g., determining whether a review is positive or negative. Using movie reviews as data, we find that standard machine learning techniques definitively outperform human-produced baselines. However, the three machine learning methods we employed (Naive Bayes, maximum entropy classification, and support vector machines) do not perform as well on sentiment classification as on traditional topic-based categorization. We conclude by examining factors that make the sentiment classification problem more challenging.

Computation and Language

Unsupervised Learning of Morphology without Morphemes

The first morphological learner based upon the theory of Whole Word Morphology Ford et al. (1997) is outlined, and preliminary evaluation results are presented. The program, Whole Word Morphologizer, takes a POS-tagged lexicon as input, induces morphological relationships without attempting to discover or identify morphemes, and is then able to generate new words beyond the learning sample. The accuracy (precision) of the generated new words is as high as 80% using the pure Whole Word theory, and 92% after a post-hoc adjustment is added to the routine.

Computation and Language

Using the Annotated Bibliography as a Resource for Indicative Summarization

We report on a language resource consisting of 2000 annotated bibliography entries, which is being analyzed as part of our research on indicative document summarization. We show how annotated bibliographies cover certain aspects of summarization that have not been well-covered by other summary corpora, and motivate why they constitute an important form to study for information retrieval. We detail our methodology for collecting the corpus, and overview our document feature markup that we introduced to facilitate summary analysis. We present the characteristics of the corpus, methods of collection, and show its use in finding the distribution of types of information included in indicative summaries and their relative ordering within the summaries.

Computation and Language

A Method for Open-Vocabulary Speech-Driven Text Retrieval

While recent retrieval techniques do not limit the number of index terms, out-of-vocabulary (OOV) words are crucial in speech recognition. Aiming at retrieving information with spoken queries, we fill the gap between speech recognition and text retrieval in terms of the vocabulary size. Given a spoken query, we generate a transcription and detect OOV words through speech recognition. We then correspond detected OOV words to terms indexed in a target collection to complete the transcription, and search the collection for documents relevant to the completed transcription. We show the effectiveness of our method by way of experiments.

Computation and Language

Japanese/English Cross-Language Information Retrieval: Exploration of Query Translation and Transliteration

Cross-language information retrieval (CLIR), where queries and documents are in different languages, has of late become one of the major topics within the information retrieval community. This paper proposes a Japanese/English CLIR system, where we combine a query translation and retrieval modules. We currently target the retrieval of technical documents, and therefore the performance of our system is highly dependent on the quality of the translation of technical terms. However, the technical term translation is still problematic in that technical terms are often compound words, and thus new terms are progressively created by combining existing base words. In addition, Japanese often represents loanwords based on its special phonogram. Consequently, existing dictionaries find it difficult to achieve sufficient coverage. To counter the first problem, we produce a Japanese/English dictionary for base words, and translate compound words on a word-by-word basis. We also use a probabilistic method to resolve translation ambiguity. For the second problem, we use a transliteration method, which corresponds words unlisted in the base word dictionary to their phonetic equivalents in the target language. We evaluate our system using a test collection for CLIR, and show that both the compound word translation and transliteration methods improve the system performance.

Computation and Language

Interleaved semantic interpretation in environment-based parsing

This paper extends a polynomial-time parsing algorithm that resolves structural ambiguity in input to a speech-based user interface by calculating and comparing the denotations of rival constituents, given some model of the interfaced application environment (Schuler 2001). The algorithm is extended to incorporate a full set of logical operators, including quantifiers and conjunctions, into this calculation without increasing the complexity of the overall algorithm beyond polynomial time, both in terms of the length of the input and the number of entities in the environment model.

Computation and Language

A Probabilistic Method for Analyzing Japanese Anaphora Integrating Zero Pronoun Detection and Resolution

This paper proposes a method to analyze Japanese anaphora, in which zero pronouns (omitted obligatory cases) are used to refer to preceding entities (antecedents). Unlike the case of general coreference resolution, zero pronouns have to be detected prior to resolution because they are not expressed in discourse. Our method integrates two probability parameters to perform zero pronoun detection and resolution in a single framework. The first parameter quantifies the degree to which a given case is a zero pronoun. The second parameter quantifies the degree to which a given entity is the antecedent for a detected zero pronoun. To compute these parameters efficiently, we use corpora with/without annotations of anaphoric relations. We show the effectiveness of our method by way of experiments.

Computation and Language

Applying a Hybrid Query Translation Method to Japanese/English Cross-Language Patent Retrieval

This paper applies an existing query translation method to cross-language patent retrieval. In our method, multiple dictionaries are used to derive all possible translations for an input query, and collocational statistics are used to resolve translation ambiguity. We used Japanese/English parallel patent abstracts to perform comparative experiments, where our method outperformed a simple dictionary-based query translation method, and achieved 76% of monolingual retrieval in terms of average precision.

Computation and Language

PRIME: A System for Multi-lingual Patent Retrieval

Given the growing number of patents filed in multiple countries, users are interested in retrieving patents across languages. We propose a multi-lingual patent retrieval system, which translates a user query into the target language, searches a multilingual database for patents relevant to the query, and improves the browsing efficiency by way of machine translation and clustering. Our system also extracts new translations from patent families consisting of comparable patents, to enhance the translation dictionary.

Computation and Language

Language Modeling for Multi-Domain Speech-Driven Text Retrieval

We report experimental results associated with speech-driven text retrieval, which facilitates retrieving information in multiple domains with spoken queries. Since users speak contents related to a target collection, we produce language models used for speech recognition based on the target collection, so as to improve both the recognition and retrieval accuracy. Experiments using existing test collections combined with dictated queries showed the effectiveness of our method.

Computation and Language

Speech-Driven Text Retrieval: Using Target IR Collections for Statistical Language Model Adaptation in Speech Recognition

Speech recognition has of late become a practical technology for real world applications. Aiming at speech-driven text retrieval, which facilitates retrieving information with spoken queries, we propose a method to integrate speech recognition and retrieval methods. Since users speak contents related to a target collection, we adapt statistical language models used for speech recognition based on the target collection, so as to improve both the recognition and retrieval accuracy. Experiments using existing test collections combined with dictated queries showed the effectiveness of our method.

Computation and Language

Using eigenvectors of the bigram graph to infer morpheme identity

This paper describes the results of some experiments exploring statistical methods to infer syntactic behavior of words and morphemes from a raw corpus in an unsupervised fashion. It shares certain points in common with Brown et al (1992) and work that has grown out of that: it employs statistical techniques to analyze syntactic behavior based on what words occur adjacent to a given word. However, we use an eigenvector decomposition of a nearest-neighbor graph to produce a two-dimensional rendering of the words of a corpus in which words of the same syntactic category tend to form neighborhoods. We exploit this technique for extending the value of automatic learning of morphology. In particular, we look at the suffixes derived from a corpus by unsupervised learning of morphology, and we ask which of these suffixes have a consistent syntactic function (e.g., in English, -tion is primarily a mark of nouns, but -s marks both noun plurals and 3rd person present on verbs), and we determine that this method works well for this task.

Computation and Language

Analysis of Titles and Readers For Title Generation Centered on the Readers

The title of a document has two roles, to give a compact summary and to lead the reader to read the document. Conventional title generation focuses on finding key expressions from the author's wording in the document to give a compact summary and pays little attention to the reader's interest. To make the title play its second role properly, it is indispensable to clarify the content (``what to say'') and wording (``how to say'') of titles that are effective to attract the target reader's interest. In this article, we first identify typical content and wording of titles aimed at general readers in a comparative study between titles of technical papers and headlines rewritten for newspapers. Next, we describe the results of a questionnaire survey on the effects of the content and wording of titles on the reader's interest. The survey of general and knowledgeable readers shows both common and different tendencies in interest.

Computation and Language

Efficient Deep Processing of Japanese

We present a broad coverage Japanese grammar written in the HPSG formalism with MRS semantics. The grammar is created for use in real world applications, such that robustness and performance issues play an important role. It is connected to a POS tagging and word segmentation tool. This grammar is being developed in a multilingual context, requiring MRS structures that are easily comparable across languages.

Computation and Language

Question Answering over Unstructured Data without Domain Restrictions

Information needs are naturally represented as questions. Automatic Natural-Language Question Answering (NLQA) has only recently become a practical task on a larger scale and without domain constraints. This paper gives a brief introduction to the field, its history and the impact of systematic evaluation competitions. It is then demonstrated that an NLQA system for English can be built and evaluated in a very short time using off-the-shelf parsers and thesauri. The system is based on Robust Minimal Recursion Semantics (RMRS) and is portable with respect to the parser used as a frontend. It applies atomic term unification supported by question classification and WordNet lookup for semantic similarity matching of parsed question representation and free text.

Computation and Language

A continuation semantics of interrogatives that accounts for Baker's ambiguity

Wh-phrases in English can appear both raised and in-situ. However, only in-situ wh-phrases can take semantic scope beyond the immediately enclosing clause. I present a denotational semantics of interrogatives that naturally accounts for these two properties. It neither invokes movement or economy, nor posits lexical ambiguity between raised and in-situ occurrences of the same wh-phrase. My analysis is based on the concept of continuations. It uses a novel type system for higher-order continuations to handle wide-scope wh-phrases while remaining strictly compositional. This treatment sheds light on the combinatorics of interrogatives as well as other kinds of so-called A'-movement.

Computation and Language

Using the DIFF Command for Natural Language Processing

Diff is a software program that detects differences between two data sets and is useful in natural language processing. This paper shows several examples of the application of diff. They include the detection of differences between two different datasets, extraction of rewriting rules, merging of two different datasets, and the optimal matching of two different data sets. Since diff comes with any standard UNIX system, it is readily available and very easy to use. Our studies showed that diff is a practical tool for research into natural language processing.

Computation and Language

Evaluation of Coreference Rules on Complex Narrative Texts

This article studies the problem of assessing relevance to each of the rules of a reference resolution system. The reference solver described here stems from a formal model of reference and is integrated in a reference processing workbench. Evaluation of the reference resolution is essential, as it enables differential evaluation of individual rules. Numerical values of these measures are given, and discussed, for simple selection rules and other processing rules; such measures are then studied for numerical parameters.

Computation and Language

Three New Methods for Evaluating Reference Resolution

Reference resolution on extended texts (several thousand references) cannot be evaluated manually. An evaluation algorithm has been proposed for the MUC tests, using equivalence classes for the coreference relation. However, we show here that this algorithm is too indulgent, yielding good scores even for poor resolution strategies. We elaborate on the same formalism to propose two new evaluation algorithms, comparing them first with the MUC algorithm and giving then results on a variety of examples. A third algorithm using only distributional comparison of equivalence classes is finally described; it assesses the relative importance of the recall vs. precision errors.

Computation and Language

Cooperation between Pronoun and Reference Resolution for Unrestricted Texts

Anaphora resolution is envisaged in this paper as part of the reference resolution process. A general open architecture is proposed, which can be particularized and configured in order to simulate some classic anaphora resolution methods. With the aim of improving pronoun resolution, the system takes advantage of elementary cues about characters of the text, which are represented through a particular data structure. In its most robust configuration, the system uses only a general lexicon, a local morpho-syntactic parser and a dictionary of synonyms. A short comparative corpus analysis shows that narrative texts are the most suitable for testing such a system.

Computation and Language

Reference Resolution Beyond Coreference: a Conceptual Frame and its Application

A model for reference use in communication is proposed, from a representationist point of view. Both the sender and the receiver of a message handle representations of their common environment, including mental representations of objects. Reference resolution by a computer is viewed as the construction of object representations using referring expressions from the discourse, whereas often only coreference links between such expressions are looked for. Differences between these two approaches are discussed. The model has been implemented with elementary rules, and tested on complex narrative texts (hundreds to thousands of referring expressions). The results support the mental representations paradigm.

Computation and Language

A Chart-Parsing Algorithm for Efficient Semantic Analysis

In some contexts, well-formed natural language cannot be expected as input to information or communication systems. In these contexts, the use of grammar-independent input (sequences of uninflected semantic units like e.g. language-independent icons) can be an answer to the users' needs. A semantic analysis can be performed, based on lexical semantic knowledge: it is equivalent to a dependency analysis with no syntactic or morphological clues. However, this requires that an intelligent system should be able to interpret this input with reasonable accuracy and in reasonable time. Here we propose a method allowing a purely semantic-based analysis of sequences of semantic units. It uses an algorithm inspired by the idea of ``chart parsing'' known in Natural Language Processing, which stores intermediate parsing results in order to bring the calculation time down. In comparison with using declarative logic programming - where the calculation time, left to a prolog engine, is hyperexponential -, this method brings the calculation time down to a polynomial time, where the order depends on the valency of the predicates.

Computation and Language

Rerendering Semantic Ontologies: Automatic Extensions to UMLS through Corpus Analytics

In this paper, we discuss the utility and deficiencies of existing ontology resources for a number of language processing applications. We describe a technique for increasing the semantic type coverage of a specific ontology, the National Library of Medicine's UMLS, with the use of robust finite state methods used in conjunction with large-scale corpus analytics of the domain corpus. We call this technique "semantic rerendering" of the ontology. This research has been done in the context of Medstract, a joint Brandeis-Tufts effort aimed at developing tools for analyzing biomedical language (i.e., Medline), as well as creating targeted databases of bio-entities, biological relations, and pathway data for biological researchers. Motivating the current research is the need to have robust and reliable semantic typing of syntactic elements in the Medline corpus, in order to improve the overall performance of the information extraction applications mentioned above.

Computation and Language

The partition semantics of questions, syntactically

Groenendijk and Stokhof (1984, 1996; Groenendijk 1999) provide a logically attractive theory of the semantics of natural language questions, commonly referred to as the partition theory. Two central notions in this theory are entailment between questions and answerhood. For example, the question "Who is going to the party?" entails the question "Is John going to the party?", and "John is going to the party" counts as an answer to both. Groenendijk and Stokhof define these two notions in terms of partitions of a set of possible worlds. We provide a syntactic characterization of entailment between questions and answerhood . We show that answers are, in some sense, exactly those formulas that are built up from instances of the question. This result lets us compare the partition theory with other approaches to interrogation -- both linguistic analyses, such as Hamblin's and Karttunen's semantics, and computational systems, such as Prolog. Our comparison separates a notion of answerhood into three aspects: equivalence (when two questions or answers are interchangeable), atomic answers (what instances of a question count as answers), and compound answers (how answers compose).

Computation and Language

Question answering: from partitions to Prolog

We implement Groenendijk and Stokhof's partition semantics of questions in a simple question answering algorithm. The algorithm is sound, complete, and based on tableau theorem proving. The algorithm relies on a syntactic characterization of answerhood: Any answer to a question is equivalent to some formula built up only from instances of the question. We prove this characterization by translating the logic of interrogation to classical predicate logic and applying Craig's interpolation theorem.

Computation and Language

Introduction to the CoNLL-2002 Shared Task: Language-Independent Named Entity Recognition

We describe the CoNLL-2002 shared task: language-independent named entity recognition. We give background information on the data sets and the evaluation method, present a general overview of the systems that have taken part in the task and discuss their performance.

Computation and Language

Probabilistic Parsing Strategies

We present new results on the relation between purely symbolic context-free parsing strategies and their probabilistic counter-parts. Such parsing strategies are seen as constructions of push-down devices from grammars. We show that preservation of probability distribution is possible under two conditions, viz. the correct-prefix property and the property of strong predictiveness. These results generalize existing results in the literature that were obtained by considering parsing strategies in isolation. From our general results we also derive negative results on so-called generalized LR parsing.

Computation and Language

Answering Subcognitive Turing Test Questions: A Reply to French

Robert French has argued that a disembodied computer is incapable of passing a Turing Test that includes subcognitive questions. Subcognitive questions are designed to probe the network of cultural and perceptual associations that humans naturally develop as we live, embodied and embedded in the world. In this paper, I show how it is possible for a disembodied computer to answer subcognitive questions appropriately, contrary to French's claim. My approach to answering subcognitive questions is to use statistical information extracted from a very large collection of text. In particular, I show how it is possible to answer a sample of subcognitive questions taken from French, by issuing queries to a search engine that indexes about 350 million Web pages. This simple algorithm may shed light on the nature of human (sub-) cognition, but the scope of this paper is limited to demonstrating that French is mistaken: a disembodied computer can answer subcognitive questions.

Computation and Language

Unsupervised Language Acquisition: Theory and Practice

In this thesis I present various algorithms for the unsupervised machine learning of aspects of natural languages using a variety of statistical models. The scientific object of the work is to examine the validity of the so-called Argument from the Poverty of the Stimulus advanced in favour of the proposition that humans have language-specific innate knowledge. I start by examining an a priori argument based on Gold's theorem, that purports to prove that natural languages cannot be learned, and some formal issues related to the choice of statistical grammars rather than symbolic grammars. I present three novel algorithms for learning various parts of natural languages: first, an algorithm for the induction of syntactic categories from unlabelled text using distributional information, that can deal with ambiguous and rare words; secondly, a set of algorithms for learning morphological processes in a variety of languages, including languages such as Arabic with non-concatenative morphology; thirdly an algorithm for the unsupervised induction of a context-free grammar from tagged text. I carefully examine the interaction between the various components, and show how these algorithms can form the basis for a empiricist model of language acquisition. I therefore conclude that the Argument from the Poverty of the Stimulus is unsupported by the evidence.

Computation and Language

An Algorithm for Aligning Sentences in Bilingual Corpora Using Lexical Information

In this paper we describe an algorithm for aligning sentences with their translations in a bilingual corpus using lexical information of the languages. Existing efficient algorithms ignore word identities and consider only the sentence lengths (Brown, 1991; Gale and Church, 1993). For a sentence in the source language text, the proposed algorithm picks the most likely translation from the target language text using lexical information and certain heuristics. It does not do statistical analysis using sentence lengths. The algorithm is language independent. It also aids in detecting addition and deletion of text in translations. The algorithm gives comparable results with the existing algorithms in most of the cases while it does better in cases where statistical algorithms do not give good results.

Computation and Language

Building an Open Language Archives Community on the OAI Foundation

The Open Language Archives Community (OLAC) is an international partnership of institutions and individuals who are creating a worldwide virtual library of language resources. The Dublin Core (DC) Element Set and the OAI Protocol have provided a solid foundation for the OLAC framework. However, we need more precision in community-specific aspects of resource description than is offered by DC. Furthermore, many of the institutions and individuals who might participate in OLAC do not have the technical resources to support the OAI protocol. This paper presents our solutions to these two problems. To address the first, we have developed an extensible application profile for language resource metadata. To address the second, we have implemented Vida (the virtual data provider) and Viser (the virtual service provider), which permit community members to provide data and services without having to implement the OAI protocol. These solutions are generic and could be adopted by other specialized subcommunities.

Computation and Language

Empirical Methods for Compound Splitting

Compounded words are a challenge for NLP applications such as machine translation (MT). We introduce methods to learn splitting rules from monolingual and parallel corpora. We evaluate them against a gold standard and measure their impact on performance of statistical MT systems. Results show accuracy of 99.1% and performance gains for MT of 0.039 BLEU on a German-English noun phrase translation task.

Computation and Language

About compression of vocabulary in computer oriented languages

The author uses the entropy of the ideal Bose-Einstein gas to minimize losses in computer-oriented languages.

Computation and Language

Glottochronology and problems of protolanguage reconstruction

A method of languages genealogical trees construction is proposed.

Computation and Language

Learning to Paraphrase: An Unsupervised Approach Using Multiple-Sequence Alignment

We address the text-to-text generation problem of sentence-level paraphrasing -- a phenomenon distinct from and more difficult than word- or phrase-level paraphrasing. Our approach applies multiple-sequence alignment to sentences gathered from unannotated comparable corpora: it learns a set of paraphrasing patterns represented by word lattice pairs and automatically determines how to apply these patterns to rewrite new sentences. The results of our evaluation experiments show that the system derives accurate paraphrases, outperforming baseline systems.

Computation and Language

Blind Normalization of Speech From Different Channels

We show how to construct a channel-independent representation of speech that has propagated through a noisy reverberant channel. This is done by blindly rescaling the cepstral time series by a non-linear function, with the form of this scale function being determined by previously encountered cepstra from that channel. The rescaled form of the time series is an invariant property of it in the following sense: it is unaffected if the time series is transformed by any time-independent invertible distortion. Because a linear channel with stationary noise and impulse response transforms cepstra in this way, the new technique can be used to remove the channel dependence of a cepstral time series. In experiments, the method achieved greater channel-independence than cepstral mean normalization, and it was comparable to the combination of cepstral mean normalization and spectral subtraction, despite the fact that no measurements of channel noise or reverberations were required (unlike spectral subtraction).

Computation and Language

Glottochronologic Retrognostic of Language System

A glottochronologic retrognostic of language system is proposed

Computation and Language

"I'm sorry Dave, I'm afraid I can't do that": Linguistics, Statistics, and Natural Language Processing circa 2001

A brief, general-audience overview of the history of natural language processing, focusing on data-driven approaches.Topics include "Ambiguity and language analysis", "Firth things first", "A 'C' change", and "The empiricists strike back".

Computation and Language

An XML based Document Suite

We report about the current state of development of a document suite and its applications. This collection of tools for the flexible and robust processing of documents in German is based on the use of XML as unifying formalism for encoding input and output data as well as process information. It is organized in modules with limited responsibilities that can easily be combined into pipelines to solve complex tasks. Strong emphasis is laid on a number of techniques to deal with lexical and conceptual gaps that are typical when starting a new application.

Computation and Language

Exploiting Sublanguage and Domain Characteristics in a Bootstrapping Approach to Lexicon and Ontology Creation

It is very costly to build up lexical resources and domain ontologies. Especially when confronted with a new application domain lexical gaps and a poor coverage of domain concepts are a problem for the successful exploitation of natural language document analysis systems that need and exploit such knowledge sources. In this paper we report about ongoing experiments with `bootstrapping techniques' for lexicon and ontology creation.

Computation and Language

An Approach for Resource Sharing in Multilingual NLP

In this paper we describe an approach for the analysis of documents in German and English with a shared pool of resources. For the analysis of German documents we use a document suite, which supports the user in tasks like information retrieval and information extraction. The core of the document suite is based on our tool XDOC. Now we want to exploit these methods for the analysis of English documents as well. For this aim we need a multilingual presentation format of the resources. These resources must be transformed into an unified format, in which we can set additional information about linguistic characteristics of the language depending on the analyzed documents. In this paper we describe our approach for such an exchange model for multilingual resources based on XML.

Computation and Language

Approximate Grammar for Information Extraction

In this paper, we present the concept of Approximate grammar and how it can be used to extract information from a documemt. As the structure of informational strings cannot be defined well in a document, we cannot use the conventional grammar rules to represent the information. Hence, the need arises to design an approximate grammar that can be used effectively to accomplish the task of Information extraction. Approximate grammars are a novel step in this direction. The rules of an approximate grammar can be given by a user or the machine can learn the rules from an annotated document. We have performed our experiments in both the above areas and the results have been impressive.

Computation and Language

Factorization of Language Models through Backing-Off Lattices

Factorization of statistical language models is the task that we resolve the most discriminative model into factored models and determine a new model by combining them so as to provide better estimate. Most of previous works mainly focus on factorizing models of sequential events, each of which allows only one factorization manner. To enable parallel factorization, which allows a model event to be resolved in more than one ways at the same time, we propose a general framework, where we adopt a backing-off lattice to reflect parallel factorizations and to define the paths along which a model is resolved into factored models, we use a mixture model to combine parallel paths in the lattice, and generalize Katz's backing-off method to integrate all the mixture models got by traversing the entire lattice. Based on this framework, we formulate two types of model factorizations that are used in natural language modeling.

Computation and Language

Techniques for effective vocabulary selection

The vocabulary of a continuous speech recognition (CSR) system is a significant factor in determining its performance. In this paper, we present three principled approaches to select the target vocabulary for a particular domain by trading off between the target out-of-vocabulary (OOV) rate and vocabulary size. We evaluate these approaches against an ad-hoc baseline strategy. Results are presented in the form of OOV rate graphs plotted against increasing vocabulary size for each technique.

Computation and Language

Bayesian Information Extraction Network

Dynamic Bayesian networks (DBNs) offer an elegant way to integrate various aspects of language in one model. Many existing algorithms developed for learning and inference in DBNs are applicable to probabilistic language modeling. To demonstrate the potential of DBNs for natural language processing, we employ a DBN in an information extraction task. We show how to assemble wealth of emerging linguistic instruments for shallow parsing, syntactic and semantic tagging, morphological decomposition, named entity recognition etc. in order to incrementally build a robust information extraction system. Our method outperforms previously published results on an established benchmark domain.

Computation and Language

The Open Language Archives Community: An infrastructure for distributed archiving of language resources

New ways of documenting and describing language via electronic media coupled with new ways of distributing the results via the World-Wide Web offer a degree of access to language resources that is unparalleled in history. At the same time, the proliferation of approaches to using these new technologies is causing serious problems relating to resource discovery and resource creation. This article describes the infrastructure that the Open Language Archives Community (OLAC) has built in order to address these problems. Its technical and usage infrastructures address problems of resource discovery by constructing a single virtual library of distributed resources. Its governance infrastructure addresses problems of resource creation by providing a mechanism through which the language-resource community can express its consensus on recommended best practices.

Computation and Language

Introduction to the CoNLL-2003 Shared Task: Language-Independent Named Entity Recognition

We describe the CoNLL-2003 shared task: language-independent named entity recognition. We give background information on the data sets (English and German) and the evaluation method, present a general overview of the systems that have taken part in the task and discuss their performance.

Computation and Language

Learning to Order Facts for Discourse Planning in Natural Language Generation

This paper presents a machine learning approach to discourse planning in natural language generation. More specifically, we address the problem of learning the most natural ordering of facts in discourse plans for a specific domain. We discuss our methodology and how it was instantiated using two different machine learning algorithms. A quantitative evaluation performed in the domain of museum exhibit descriptions indicates that our approach performs significantly better than manually constructed ordering rules. Being retrainable, the resulting planners can be ported easily to other similar domains, without requiring language technology expertise.

Computation and Language

An Improved k-Nearest Neighbor Algorithm for Text Categorization

k is the most important parameter in a text categorization system based on k-Nearest Neighbor algorithm (kNN).In the classification process, k nearest documents to the test one in the training set are determined firstly. Then, the predication can be made according to the category distribution among these k nearest neighbors. Generally speaking, the class distribution in the training set is uneven. Some classes may have more samples than others. Therefore, the system performance is very sensitive to the choice of the parameter k. And it is very likely that a fixed k value will result in a bias on large categories. To deal with these problems, we propose an improved kNN algorithm, which uses different numbers of nearest neighbors for different categories, rather than a fixed number across all categories. More samples (nearest neighbors) will be used for deciding whether a test document should be classified to a category, which has more samples in the training set. Preliminary experiments on Chinese text categorization show that our method is less sensitive to the parameter k than the traditional one, and it can properly classify documents belonging to smaller classes with a large k. The method is promising for some cases, where estimating the parameter k via cross-validation is not allowed.

Computation and Language

Anusaaraka: Machine Translation in Stages

Fully-automatic general-purpose high-quality machine translation systems (FGH-MT) are extremely difficult to build. In fact, there is no system in the world for any pair of languages which qualifies to be called FGH-MT. The reasons are not far to seek. Translation is a creative process which involves interpretation of the given text by the translator. Translation would also vary depending on the audience and the purpose for which it is meant. This would explain the difficulty of building a machine translation system. Since, the machine is not capable of interpreting a general text with sufficient accuracy automatically at present - let alone re-expressing it for a given audience, it fails to perform as FGH-MT. FOOTNOTE{The major difficulty that the machine faces in interpreting a given text is the lack of general world knowledge or common sense knowledge.}

Computation and Language

Issues in Communication Game

As interaction between autonomous agents, communication can be analyzed in game-theoretic terms. Meaning game is proposed to formalize the core of intended communication in which the sender sends a message and the receiver attempts to infer its meaning intended by the sender. Basic issues involved in the game of natural language communication are discussed, such as salience, grammaticality, common sense, and common belief, together with some demonstration of the feasibility of game-theoretic account of language.

Computation and Language

Parsing and Generation with Tabulation and Compilation

The standard tabulation techniques for logic programming presuppose fixed order of computation. Some data-driven control should be introduced in order to deal with diverse contexts. The present paper describes a data-driven method of constraint transformation with a sort of compilation which subsumes accessibility check and last-call optimization, which characterize standard natural-language parsing techniques, semantic-head-driven generation, etc.

Computation and Language

The Linguistic DS: Linguisitic Description in MPEG-7

MPEG-7 (Moving Picture Experts Group Phase 7) is an XML-based international standard on semantic description of multimedia content. This document discusses the Linguistic DS and related tools. The linguistic DS is a tool, based on the GDA tag set (http://i-content.org/GDA/tagset.html), for semantic annotation of linguistic data in or associated with multimedia content. The current document text reflects `Study of FPDAM - MPEG-7 MDS Extensions' issued in March 2003, and not most part of MPEG-7 MDS, for which the readers are referred to the first version of MPEG-7 MDS document available from ISO (http://www.iso.org). Without that reference, however, this document should be mostly intelligible to those who are familiar with XML and linguistic theories. Comments are welcome and will be considered in the standardization process.

Computation and Language

Collaborative Creation of Digital Content in Indian Languages

The world is passing through a major revolution called the information revolution, in which information and knowledge is becoming available to people in unprecedented amounts wherever and whenever they need it. Those societies which fail to take advantage of the new technology will be left behind, just like in the industrial revolution. The information revolution is based on two major technologies: computers and communication. These technologies have to be delivered in a COST EFFECTIVE manner, and in LANGUAGES accessible to people. One way to deliver them in cost effective manner is to make suitable technology choices, and to allow people to access through shared resources. This could be done throuch street corner shops (for computer usage, e-mail etc.), schools, community centres and local library centres.

Computation and Language

Information Revolution

The world is passing through a major revolution called the information revolution, in which information and knowledge is becoming available to people in unprecedented amounts wherever and whenever they need it. Those societies which fail to take advantage of the new technology will be left behind, just like in the industrial revolution. The information revolution is based on two major technologies: computers and communication. These technologies have to be delivered in a COST EFFECTIVE manner, and in LANGUAGES accessible to people. One way to deliver them in cost effective manner is to make suitable technology choices (discussed later), and to allow people to access through shared resources. This could be done throuch street corner shops (for computer usage, e-mail etc.), schools, community centers and local library centres.

Computation and Language

Anusaaraka: Overcoming the Language Barrier in India

The anusaaraka system makes text in one Indian language accessible in another Indian language. In the anusaaraka approach, the load is so divided between man and computer that the language load is taken by the machine, and the interpretation of the text is left to the man. The machine presents an image of the source text in a language close to the target language.In the image, some constructions of the source language (which do not have equivalents) spill over to the output. Some special notation is also devised. The user after some training learns to read and understand the output. Because the Indian languages are close, the learning time of the output language is short, and is expected to be around 2 weeks. The output can also be post-edited by a trained user to make it grammatically correct in the target language. Style can also be changed, if necessary. Thus, in this scenario, it can function as a human assisted translation system. Currently, anusaarakas are being built from Telugu, Kannada, Marathi, Bengali and Punjabi to Hindi. They can be built for all Indian languages in the near future. Everybody must pitch in to build such systems connecting all Indian languages, using the free software model.

Computation and Language

Language Access: An Information Based Approach

The anusaaraka system (a kind of machine translation system) makes text in one Indian language accessible through another Indian language. The machine presents an image of the source text in a language close to the target language. In the image, some constructions of the source language (which do not have equivalents in the target language) spill over to the output. Some special notation is also devised. Anusaarakas have been built from five pairs of languages: Telugu,Kannada, Marathi, Bengali and Punjabi to Hindi. They are available for use through Email servers. Anusaarkas follows the principle of substitutibility and reversibility of strings produced. This implies preservation of information while going from a source language to a target language. For narrow subject areas, specialized modules can be built by putting subject domain knowledge into the system, which produce good quality grammatical output. However, it should be remembered, that such modules will work only in narrow areas, and will sometimes go wrong. In such a situation, anusaaraka output will still remain useful.

Computation and Language

LERIL : Collaborative Effort for Creating Lexical Resources

The paper reports on efforts taken to create lexical resources pertaining to Indian languages, using the collaborative model. The lexical resources being developed are: (1) Transfer lexicon and grammar from English to several Indian languages. (2) Dependencey tree bank of annotated corpora for several Indian languages. The dependency trees are based on the Paninian model. (3) Bilingual dictionary of 'core meanings'.

Computation and Language

Extending Dublin Core Metadata to Support the Description and Discovery of Language Resources

As language data and associated technologies proliferate and as the language resources community expands, it is becoming increasingly difficult to locate and reuse existing resources. Are there any lexical resources for such-and-such a language? What tool works with transcripts in this particular format? What is a good format to use for linguistic data of this type? Questions like these dominate many mailing lists, since web search engines are an unreliable way to find language resources. This paper reports on a new digital infrastructure for discovering language resources being developed by the Open Language Archives Community (OLAC). At the core of OLAC is its metadata format, which is designed to facilitate description and discovery of all kinds of language resources, including data, tools, or advice. The paper describes OLAC metadata, its relationship to Dublin Core metadata, and its dissemination using the metadata harvesting protocol of the Open Archives Initiative.

Computation and Language

Evaluation of text data mining for database curation: lessons learned from the KDD Challenge Cup

MOTIVATION: The biological literature is a major repository of knowledge. Many biological databases draw much of their content from a careful curation of this literature. However, as the volume of literature increases, the burden of curation increases. Text mining may provide useful tools to assist in the curation process. To date, the lack of standards has made it impossible to determine whether text mining techniques are sufficiently mature to be useful. RESULTS: We report on a Challenge Evaluation task that we created for the Knowledge Discovery and Data Mining (KDD) Challenge Cup. We provided a training corpus of 862 articles consisting of journal articles curated in FlyBase, along with the associated lists of genes and gene products, as well as the relevant data fields from FlyBase. For the test, we provided a corpus of 213 new (`blind') articles; the 18 participating groups provided systems that flagged articles for curation, based on whether the article contained experimental evidence for gene expression products. We report on the the evaluation results and describe the techniques used by the top performing groups. CONTACT: asy@mitre.org KEYWORDS: text mining, evaluation, curation, genomics, data management

Computation and Language

Building a Test Collection for Speech-Driven Web Retrieval

This paper describes a test collection (benchmark data) for retrieval systems driven by spoken queries. This collection was produced in the subtask of the NTCIR-3 Web retrieval task, which was performed in a TREC-style evaluation workshop. The search topics and document collection for the Web retrieval task were used to produce spoken queries and language models for speech recognition, respectively. We used this collection to evaluate the performance of our retrieval system. Experimental results showed that (a) the use of target documents for language modeling and (b) enhancement of the vocabulary size in speech recognition were effective in improving the system performance.

Computation and Language

A Cross-media Retrieval System for Lecture Videos

We propose a cross-media lecture-on-demand system, in which users can selectively view specific segments of lecture videos by submitting text queries. Users can easily formulate queries by using the textbook associated with a target lecture, even if they cannot come up with effective keywords. Our system extracts the audio track from a target lecture video, generates a transcription by large vocabulary continuous speech recognition, and produces a text index. Experimental results showed that by adapting speech recognition to the topic of the lecture, the recognition accuracy increased and the retrieval accuracy was comparable with that obtained by human transcription.

Computation and Language

Measuring Praise and Criticism: Inference of Semantic Orientation from Association

The evaluative character of a word is called its semantic orientation. Positive semantic orientation indicates praise (e.g., "honest", "intrepid") and negative semantic orientation indicates criticism (e.g., "disturbing", "superfluous"). Semantic orientation varies in both direction (positive or negative) and degree (mild to strong). An automated system for measuring semantic orientation would have application in text classification, text filtering, tracking opinions in online discussions, analysis of survey responses, and automated chat systems (chatbots). This paper introduces a method for inferring the semantic orientation of a word from its statistical association with a set of positive and negative paradigm words. Two instances of this approach are evaluated, based on two different statistical measures of word association: pointwise mutual information (PMI) and latent semantic analysis (LSA). The method is experimentally tested with 3,596 words (including adjectives, adverbs, nouns, and verbs) that have been manually labeled positive (1,614 words) and negative (1,982 words). The method attains an accuracy of 82.8% on the full test set, but the accuracy rises above 95% when the algorithm is allowed to abstain from classifying mild words.

Computation and Language

Combining Independent Modules to Solve Multiple-choice Synonym and Analogy Problems

Existing statistical approaches to natural language problems are very coarse approximations to the true complexity of language processing. As such, no single technique will be best for all problem instances. Many researchers are examining ensemble methods that combine the output of successful, separately developed modules to create more accurate solutions. This paper examines three merging rules for combining probability distributions: the well known mixture rule, the logarithmic rule, and a novel product rule. These rules were applied with state-of-the-art results to two problems commonly used to assess human mastery of lexical semantics -- synonym questions and analogy questions. All three merging rules result in ensembles that are more accurate than any of their component modules. The differences among the three rules are not statistically significant, but it is suggestive that the popular mixture rule is not the best rule for either of the two problems.

Computation and Language

Effective XML Representation for Spoken Language in Organisations

Spoken Language can be used to provide insights into organisational processes, unfortunately transcription and coding stages are very time consuming and expensive. The concept of partial transcription and coding is proposed in which spoken language is indexed prior to any subsequent processing. The functional linguistic theory of texture is used to describe the effects of partial transcription on observational records. The standard used to encode transcript context and metadata is called CHAT, but a previous XML schema developed to implement it contains design assumptions that make it difficult to support partial transcription for example. This paper describes a more effective XML schema that overcomes many of these problems and is intended for use in applications that support the rapid development of spoken language deliverables.

Computation and Language

A Dynamic Programming Algorithm for the Segmentation of Greek Texts

In this paper we introduce a dynamic programming algorithm to perform linear text segmentation by global minimization of a segmentation cost function which consists of: (a) within-segment word similarity and (b) prior information about segment length. The evaluation of the segmentation accuracy of the algorithm on a text collection consisting of Greek texts showed that the algorithm achieves high segmentation accuracy and appears to be very innovating and promissing.

Computation and Language

Application Architecture for Spoken Language Resources in Organisational Settings

Special technologies need to be used to take advantage of, and overcome, the challenges associated with acquiring, transforming, storing, processing, and distributing spoken language resources in organisations. This paper introduces an application architecture consisting of tools and supporting utilities for indexing and transcription, and describes how these tools, together with downstream processing and distribution systems, can be integrated into a workflow. Two sample applications for this architecture are outlined- the analysis of decision-making processes in organisations and the deployment of systems development methods by designers in the field.

Computation and Language

The Rank-Frequency Analysis for the Functional Style Corpora in the Ukrainian Language

We use the rank-frequency analysis for the estimation of Kernel Vocabulary size within specific corpora of Ukrainian. The extrapolation of high-rank behaviour is utilized for estimation of the total vocabulary size.

Computation and Language

Measuring the Functional Load of Phonological Contrasts

Frequency counts are a measure of how much use a language makes of a linguistic unit, such as a phoneme or word. However, what is often important is not the units themselves, but the contrasts between them. A measure is therefore needed for how much use a language makes of a contrast, i.e. the functional load (FL) of the contrast. We generalize previous work in linguistics and speech recognition and propose a family of measures for the FL of several phonological contrasts, including phonemic oppositions, distinctive features, suprasegmentals, and phonological rules. We then test it for robustness to changes of corpora. Finally, we provide examples in Cantonese, Dutch, English, German and Mandarin, in the context of historical linguistics, language acquisition and speech recognition. More information can be found at http://dinoj.info/research/fload

Computation and Language

Embedding Web-based Statistical Translation Models in Cross-Language Information Retrieval

Although more and more language pairs are covered by machine translation services, there are still many pairs that lack translation resources. Cross-language information retrieval (CLIR) is an application which needs translation functionality of a relatively low level of sophistication since current models for information retrieval (IR) are still based on a bag-of-words. The Web provides a vast resource for the automatic construction of parallel corpora which can be used to train statistical translation models automatically. The resulting translation models can be embedded in several ways in a retrieval model. In this paper, we will investigate the problem of automatically mining parallel texts from the Web and different ways of integrating the translation models within the retrieval process. Our experiments on standard test collections for CLIR show that the Web-based translation models can surpass commercial MT systems in CLIR tasks. These results open the perspective of constructing a fully automatic query translation device for CLIR at a very low cost.

Computation and Language

A Flexible Pragmatics-driven Language Generator for Animated Agents

This paper describes the NECA MNLG; a fully implemented Multimodal Natural Language Generation module. The MNLG is deployed as part of the NECA system which generates dialogues between animated agents. The generation module supports the seamless integration of full grammar rules, templates and canned text. The generator takes input which allows for the specification of syntactic, semantic and pragmatic constraints on the output.

Computation and Language

Towards Automated Generation of Scripted Dialogue: Some Time-Honoured Strategies

The main aim of this paper is to introduce automated generation of scripted dialogue as a worthwhile topic of investigation. In particular the fact that scripted dialogue involves two layers of communication, i.e., uni-directional communication between the author and the audience of a scripted dialogue and bi-directional pretended communication between the characters featuring in the dialogue, is argued to raise some interesting issues. Our hope is that the combined study of the two layers will forge links between research in text generation and dialogue processing. The paper presents a first attempt at creating such links by studying three types of strategies for the automated generation of scripted dialogue. The strategies are derived from examples of human-authored and naturally occurring dialogue.

Computation and Language

Dialogue as Discourse: Controlling Global Properties of Scripted Dialogue

This paper explains why scripted dialogue shares some crucial properties with discourse. In particular, when scripted dialogues are generated by a Natural Language Generation system, the generator can apply revision strategies that cannot normally be used when the dialogue results from an interaction between autonomous agents (i.e., when the dialogue is not scripted). The paper explains that the relevant revision operators are best applied at the level of a dialogue plan and discusses how the generator may decide when to apply a given revision operator.

Computation and Language

Acquiring Lexical Paraphrases from a Single Corpus

This paper studies the potential of identifying lexical paraphrases within a single corpus, focusing on the extraction of verb paraphrases. Most previous approaches detect individual paraphrase instances within a pair (or set) of comparable corpora, each of them containing roughly the same information, and rely on the substantial level of correspondence of such corpora. We present a novel method that successfully detects isolated paraphrase instances within a single corpus without relying on any a-priori structure and information. A comparison suggests that an instance-based approach may be combined with a vector based approach in order to assess better the paraphrase likelihood for many verb pairs.

Computation and Language

Part-of-Speech Tagging with Minimal Lexicalization

We use a Dynamic Bayesian Network to represent compactly a variety of sublexical and contextual features relevant to Part-of-Speech (PoS) tagging. The outcome is a flexible tagger (LegoTag) with state-of-the-art performance (3.6% error on a benchmark corpus). We explore the effect of eliminating redundancy and radically reducing the size of feature vocabularies. We find that a small but linguistically motivated set of suffixes results in improved cross-corpora generalization. We also show that a minimal lexicon limited to function words is sufficient to ensure reasonable performance.

Computation and Language

Lexical Base as a Compressed Language Model of the World (on the material of the Ukrainian language)

In the article the fact is verified that the list of words selected by formal statistical methods (frequency and functional genre unrestrictedness) is not a conglomerate of non-related words. It creates a system of interrelated items and it can be named "lexical base of language". This selected list of words covers all the spheres of human activities. To verify this statement the invariant synoptical scheme common for ideographic dictionaries of different language was determined.

Computation and Language

A Flexible Rule Compiler for Speech Synthesis

We present a flexible rule compiler developed for a text-to-speech (TTS) system. The compiler converts a set of rules into a finite-state transducer (FST). The input and output of the FST are subject to parameterization, so that the system can be applied to strings and sequences of feature-structures. The resulting transducer is guaranteed to realize a function (as opposed to a relation), and therefore can be implemented as a deterministic device (either a deterministic FST or a bimachine).

Computation and Language

Delimited continuations in natural language: quantification and polarity sensitivity

Making a linguistic theory is like making a programming language: one typically devises a type system to delineate the acceptable utterances and a denotational semantics to explain observations on their behavior. Via this connection, the programming language concept of delimited continuations can help analyze natural language phenomena such as quantification and polarity sensitivity. Using a logical metalanguage whose syntax includes control operators and whose semantics involves evaluation order, these analyses can be expressed in direct style rather than continuation-passing style, and these phenomena can be thought of as computational side effects.

Computation and Language

Polarity sensitivity and evaluation order in type-logical grammar

We present a novel, type-logical analysis of_polarity sensitivity_: how negative polarity items (like "any" and "ever") or positive ones (like "some") are licensed or prohibited. It takes not just scopal relations but also linear order into account, using the programming-language notions of delimited continuations and evaluation order, respectively. It thus achieves greater empirical coverage than previous proposals.

Computation and Language

Tabular Parsing

This is a tutorial on tabular parsing, on the basis of tabulation of nondeterministic push-down automata. Discussed are Earley's algorithm, the Cocke-Kasami-Younger algorithm, tabular LR parsing, the construction of parse trees, and further issues.

Computation and Language

NLML--a Markup Language to Describe the Unlimited English Grammar

In this paper we present NLML (Natural Language Markup Language), a markup language to describe the syntactic and semantic structure of any grammatically correct English expression. At first the related works are analyzed to demonstrate the necessity of the NLML: simple form, easy management and direct storage. Then the description of the English grammar with NLML is introduced in details in three levels: sentences (with different complexities, voices, moods, and tenses), clause (relative clause and noun clause) and phrase (noun phrase, verb phrase, prepositional phrase, adjective phrase, adverb phrase and predicate phrase). At last the application fields of the NLML in NLP are shown with two typical examples: NLOJM (Natural Language Object Modal in Java) and NLDB (Natural Language Database).

Computation and Language

Test Collections for Patent-to-Patent Retrieval and Patent Map Generation in NTCIR-4 Workshop

This paper describes the Patent Retrieval Task in the Fourth NTCIR Workshop, and the test collections produced in this task. We perform the invalidity search task, in which each participant group searches a patent collection for the patents that can invalidate the demand in an existing claim. We also perform the automatic patent map generation task, in which the patents associated with a specific topic are organized in a multi-dimensional matrix.

Computation and Language

NLOMJ--Natural Language Object Model in Java

In this paper we present NLOMJ--a natural language object model in Java with English as the experiment language. This modal describes the grammar elements of any permissible expression in a natural language and their complicated relations with each other with the concept "Object" in OOP(Object Oriented Programming). Directly mapped to the syntax and semantics of the natural language, it can be used in information retrieval as a linguistic method. Around the UML diagram of the NLOMJ the important classes(Sentence, Clause and Phrase) and their sub classes are introduced and their syntactic and semantic meanings are explained.

Computation and Language

Exploiting Cross-Document Relations for Multi-document Evolving Summarization

This paper presents a methodology for summarization from multiple documents which are about a specific topic. It is based on the specification and identification of the cross-document relations that occur among textual elements within those documents. Our methodology involves the specification of the topic-specific entities, the messages conveyed for the specific entities by certain textual elements and the specification of the relations that can hold among these messages. The above resources are necessary for setting up a specific topic for our query-based summarization approach which uses these resources to identify the query-specific messages within the documents and the query-specific relations that connect these messages across documents.

Computation and Language

A Probabilistic Model of Machine Translation

A probabilistic model for computer-based generation of a machine translation system on the basis of English-Russian parallel text corpora is suggested. The model is trained using parallel text corpora with pre-aligned source and target sentences. The training of the model results in a bilingual dictionary of words and "word blocks" with relevant translation probability.

Computation and Language

Catching the Drift: Probabilistic Content Models, with Applications to Generation and Summarization

We consider the problem of modeling the content structure of texts within a specific domain, in terms of the topics the texts address and the order in which these topics appear. We first present an effective knowledge-lean method for learning content models from un-annotated documents, utilizing a novel adaptation of algorithms for Hidden Markov Models. We then apply our method to two complementary tasks: information ordering and extractive summarization. Our experiments show that incorporating content models in these applications yields substantial improvement over previously-proposed methods.

Computation and Language

Algorithms for weighted multi-tape automata

This report defines various operations for weighted multi-tape automata (WMTAs) and describes algorithms that have been implemented for those operations in the WFSC toolkit. Some algorithms are new, others are known or similar to known algorithms. The latter will be recalled to make this report more complete and self-standing. We present a new approach to multi-tape intersection, meaning the intersection of a number of tapes of one WMTA with the same number of tapes of another WMTA. In our approach, multi-tape intersection is not considered as an atomic operation but rather as a sequence of more elementary ones, which facilitates its implementation. We show an example of multi-tape intersection, actually transducer intersection, that can be compiled with our approach but not with several other methods that we analysed. To show the practical relavance of our work, we include an example of application: the preservation of intermediate results in transduction cascades.

Computation and Language

Zipf's law and the creation of musical context

This article discusses the extension of the notion of context from linguistics to the domain of music. In language, the statistical regularity known as Zipf's law -which concerns the frequency of usage of different words- has been quantitatively related to the process of text generation. This connection is established by Simon's model, on the basis of a few assumptions regarding the accompanying creation of context. Here, it is shown that the statistics of note usage in musical compositions are compatible with the predictions of Simon's model. This result, which gives objective support to the conceptual likeness of context in language and music, is obtained through automatic analysis of the digital versions of several compositions. As a by-product, a quantitative measure of context definiteness is introduced and used to compare tonal and atonal works.

Computation and Language

A Public Reference Implementation of the RAP Anaphora Resolution Algorithm

This paper describes a standalone, publicly-available implementation of the Resolution of Anaphora Procedure (RAP) given by Lappin and Leass (1994). The RAP algorithm resolves third person pronouns, lexical anaphors, and identifies pleonastic pronouns. Our implementation, JavaRAP, fills a current need in anaphora resolution research by providing a reference implementation that can be benchmarked against current algorithms. The implementation uses the standard, publicly available Charniak (2000) parser as input, and generates a list of anaphora-antecedent pairs as output. Alternately, an in-place annotation or substitution of the anaphors with their antecedents can be produced. Evaluation on the MUC-6 co-reference task shows that JavaRAP has an accuracy of 57.9%, similar to the performance given previously in the literature (e.g., Preiss 2002).

Computation and Language

Building a linguistic corpus from bee dance data

This paper discusses the problems and possibility of collecting bee dance data in a linguistic \textit{corpus} and use linguistic instruments such as Zipf's law and entropy statistics to decide on the question whether the dance carries information of any kind. We describe this against the historical background of attempts to analyse non-human communication systems.

Computation and Language

Annotating Predicate-Argument Structure for a Parallel Treebank

We report on a recently initiated project which aims at building a multi-layered parallel treebank of English and German. Particular attention is devoted to a dedicated predicate-argument layer which is used for aligning translationally equivalent sentences of the two languages. We describe both our conceptual decisions and aspects of their technical realisation. We discuss some selected problems and conclude with a few remarks on how this project relates to similar projects in the field.

Computation and Language

Statistical Machine Translation by Generalized Parsing

Designers of statistical machine translation (SMT) systems have begun to employ tree-structured translation models. Systems involving tree-structured translation models tend to be complex. This article aims to reduce the conceptual complexity of such systems, in order to make them easier to design, implement, debug, use, study, understand, explain, modify, and improve. In service of this goal, the article extends the theory of semiring parsing to arrive at a novel abstract parsing algorithm with five functional parameters: a logic, a grammar, a semiring, a search strategy, and a termination condition. The article then shows that all the common algorithms that revolve around tree-structured translation models, including hierarchical alignment, inference for parameter estimation, translation, and structured evaluation, can be derived by generalizing two of these parameters -- the grammar and the logic. The article culminates with a recipe for using such generalized parsers to train, apply, and evaluate an SMT system that is driven by tree-structured translation models.

Computation and Language

Summarizing Encyclopedic Term Descriptions on the Web

We are developing an automatic method to compile an encyclopedic corpus from the Web. In our previous work, paragraph-style descriptions for a term are extracted from Web pages and organized based on domains. However, these descriptions are independent and do not comprise a condensed text as in hand-crafted encyclopedias. To resolve this problem, we propose a summarization method, which produces a single text from multiple descriptions. The resultant summary concisely describes a term from different viewpoints. We also show the effectiveness of our method by means of experiments.

Computation and Language

Unsupervised Topic Adaptation for Lecture Speech Retrieval

We are developing a cross-media information retrieval system, in which users can view specific segments of lecture videos by submitting text queries. To produce a text index, the audio track is extracted from a lecture video and a transcription is generated by automatic speech recognition. In this paper, to improve the quality of our retrieval system, we extensively investigate the effects of adapting acoustic and language models on speech recognition. We perform an MLLR-based method to adapt an acoustic model. To obtain a corpus for language model adaptation, we use the textbook for a target lecture to search a Web collection for the pages associated with the lecture topic. We show the effectiveness of our method by means of experiments.

Computation and Language

Effects of Language Modeling on Speech-driven Question Answering

We integrate automatic speech recognition (ASR) and question answering (QA) to realize a speech-driven QA system, and evaluate its performance. We adapt an N-gram language model to natural language questions, so that the input of our system can be recognized with a high accuracy. We target WH-questions which consist of the topic part and fixed phrase used to ask about something. We first produce a general N-gram model intended to recognize the topic and emphasize the counts of the N-grams that correspond to the fixed phrases. Given a transcription by the ASR engine, the QA engine extracts the answer candidates from target documents. We propose a passage retrieval method robust against recognition errors in the transcription. We use the QA test collection produced in NTCIR, which is a TREC-style evaluation workshop, and show the effectiveness of our method by means of experiments.

Computation and Language

A Bimachine Compiler for Ranked Tagging Rules

This paper describes a novel method of compiling ranked tagging rules into a deterministic finite-state device called a bimachine. The rules are formulated in the framework of regular rewrite operations and allow unrestricted regular expressions in both left and right rule contexts. The compiler is illustrated by an application within a speech synthesis system.

Computation and Language

Word Sense Disambiguation by Web Mining for Word Co-occurrence Probabilities

This paper describes the National Research Council (NRC) Word Sense Disambiguation (WSD) system, as applied to the English Lexical Sample (ELS) task in Senseval-3. The NRC system approaches WSD as a classical supervised machine learning problem, using familiar tools such as the Weka machine learning software and Brill's rule-based part-of-speech tagger. Head words are represented as feature vectors with several hundred features. Approximately half of the features are syntactic and the other half are semantic. The main novelty in the system is the method for generating the semantic features, based on word \hbox{co-occurrence} probabilities. The probabilities are estimated using the Waterloo MultiText System with a corpus of about one terabyte of unlabeled text, collected by a web crawler.

Computation and Language

Incremental Construction of Minimal Acyclic Sequential Transducers from Unsorted Data

This paper presents an efficient algorithm for the incremental construction of a minimal acyclic sequential transducer (ST) for a dictionary consisting of a list of input and output strings. The algorithm generalises a known method of constructing minimal finite-state automata (Daciuk et al. 2000). Unlike the algorithm published by Mihov and Maurel (2001), it does not require the input strings to be sorted. The new method is illustrated by an application to pronunciation dictionaries.

Computation and Language