id
stringlengths
9
16
title
stringlengths
4
278
categories
stringlengths
5
104
abstract
stringlengths
6
4.09k
cmp-lg/9406021
A symbolic description of punning riddles and its computer implementation
cmp-lg cs.CL
Riddles based on simple puns can be classified according to the patterns of word, syllable or phrase similarity they depend upon. We have devised a formal model of the semantic and syntactic regularities underlying some of the simpler types of punning riddle. We have also implemented this preliminary theory in a computer program which can generate riddles from a lexicon containing general data about words and phrases; that is, the lexicon content is not customised to produce jokes. Informal evaluation of the program's results by a set of human judges suggest that the riddles produced by this program are of comparable quality to those in general circulation among school children.
cmp-lg/9406022
An implemented model of punning riddles
cmp-lg cs.CL
In this paper, we discuss a model of simple question-answer punning, implemented in a program, JAPE, which generates riddles from humour-independent lexical entries. The model uses two main types of structure: schemata, which determine the relationships between key words in a joke, and templates, which produce the surface form of the joke. JAPE succeeds in generating pieces of text that are recognizably jokes, but some of them are not very good jokes. We mention some potential improvements and extensions, including post-production heuristics for ordering the jokes according to quality.
cmp-lg/9406023
A Spanish Tagset for the CRATER Project
cmp-lg cs.CL
This working paper describes the Spanish tagset to be used in the context of CRATER, a CEC funded project aiming at the creation of a multilingual (English, French, Spanish) aligned corpus using the International Telecommunications Union corpus. In this respect, each version of the corpus will be (or is currently) tagged. Xerox PARC tagger will be adapted to Spanish in order to perform the tagging of the Spanish version. This tagset has been devised as the ideal one for Spanish, and has been posted to several lists in order to get feedback to it.
cmp-lg/9406024
Learning Fault-tolerant Speech Parsing with SCREEN
cmp-lg cs.CL
This paper describes a new approach and a system SCREEN for fault-tolerant speech parsing. SCREEEN stands for Symbolic Connectionist Robust EnterprisE for Natural language. Speech parsing describes the syntactic and semantic analysis of spontaneous spoken language. The general approach is based on incremental immediate flat analysis, learning of syntactic and semantic speech parsing, parallel integration of current hypotheses, and the consideration of various forms of speech related errors. The goal for this approach is to explore the parallel interactions between various knowledge sources for learning incremental fault-tolerant speech parsing. This approach is examined in a system SCREEN using various hybrid connectionist techniques. Hybrid connectionist techniques are examined because of their promising properties of inherent fault tolerance, learning, gradedness and parallel constraint integration. The input for SCREEN is hypotheses about recognized words of a spoken utterance potentially analyzed by a speech system, the output is hypotheses about the flat syntactic and semantic analysis of the utterance. In this paper we focus on the general approach, the overall architecture, and examples for learning flat syntactic speech parsing. Different from most other speech language architectures SCREEN emphasizes an interactive rather than an autonomous position, learning rather than encoding, flat analysis rather than in-depth analysis, and fault-tolerant processing of phonetic, syntactic and semantic knowledge.
cmp-lg/9406025
Emergent Parsing and Generation with Generalized Chart
cmp-lg cs.CL
A new, flexible inference method for Horn logic program is proposed, which is a drastic generalization of chart parsing, partial instantiations of clauses in a program roughly corresponding to arcs in a chart. Chart-like parsing and semantic-head-driven generation emerge from this method. With a parsimonious instantiation scheme for ambiguity packing, the parsing complexity reduces to that of standard chart-based algorithms.
cmp-lg/9406026
The Very Idea of Dynamic Semantics
cmp-lg cs.CL
"Natural languages are programming languages for minds." Can we or should we take this slogan seriously? If so, how? Can answers be found by looking at the various "dynamic" treatments of natural language developed over the last decade or so, mostly in response to problems associated with donkey anaphora? In Dynamic Logic of Programs, the meaning of a program is a binary relation on the set of states of some abstract machine. This relation is meant to model aspects of the effects of the execution of the program, in particular its input-output behavior. What, if anything, are the dynamic aspects of various proposed dynamic semantics for natural languages supposed to model? Is there anything dynamic to be modeled? If not, what is all the full about? We shall try to answer some, at least, of these questions and provide materials for answers to others.
cmp-lg/9406027
Analyzing and Improving Statistical Language Models for Speech Recognition
cmp-lg cs.CL
In many current speech recognizers, a statistical language model is used to indicate how likely it is that a certain word will be spoken next, given the words recognized so far. How can statistical language models be improved so that more complex speech recognition tasks can be tackled? Since the knowledge of the weaknesses of any theory often makes improving the theory easier, the central idea of this thesis is to analyze the weaknesses of existing statistical language models in order to subsequently improve them. To that end, we formally define a weakness of a statistical language model in terms of the logarithm of the total probability, LTP, a term closely related to the standard perplexity measure used to evaluate statistical language models. We apply our definition of a weakness to a frequently used statistical language model, called a bi-pos model. This results, for example, in a new modeling of unknown words which improves the performance of the model by 14% to 21%. Moreover, one of the identified weaknesses has prompted the development of our generalized N-pos language model, which is also outlined in this thesis. It can incorporate linguistic knowledge even if it extends over many words and this is not feasible in a traditional N-pos model. This leads to a discussion of whatknowledge should be added to statistical language models in general and we give criteria for selecting potentially useful knowledge. These results show the usefulness of both our definition of a weakness and of performing an analysis of weaknesses of statistical language models in general.
cmp-lg/9406028
Resolution of Syntactic Ambiguity: the Case of New Subjects
cmp-lg cs.CL
I review evidence for the claim that syntactic ambiguities are resolved on the basis of the meaning of the competing analyses, not their structure. I identify a collection of ambiguities that do not yet have a meaning-based account and propose one which is based on the interaction of discourse and grammatical function. I provide evidence for my proposal by examining statistical properties of the Penn Treebank of syntactically annotated text.
cmp-lg/9406029
A Computational Model of Syntactic Processing: Ambiguity Resolution from Interpretation
cmp-lg cs.CL
Syntactic ambiguity abounds in natural language, yet humans have no difficulty coping with it. In fact, the process of ambiguity resolution is almost always unconscious. But it is not infallible, however, as example 1 demonstrates. 1. The horse raced past the barn fell. This sentence is perfectly grammatical, as is evident when it appears in the following context: 2. Two horses were being shown off to a prospective buyer. One was raced past a meadow. and the other was raced past a barn. ... Grammatical yet unprocessable sentences such as 1 are called `garden-path sentences.' Their existence provides an opportunity to investigate the human sentence processing mechanism by studying how and when it fails. The aim of this thesis is to construct a computational model of language understanding which can predict processing difficulty. The data to be modeled are known examples of garden path and non-garden path sentences, and other results from psycholinguistics. It is widely believed that there are two distinct loci of computation in sentence processing: syntactic parsing and semantic interpretation. One longstanding controversy is which of these two modules bears responsibility for the immediate resolution of ambiguity. My claim is that it is the latter, and that the syntactic processing module is a very simple device which blindly and faithfully constructs all possible analyses for the sentence up to the current point of processing. The interpretive module serves as a filter, occasionally discarding certain of these analyses which it deems less appropriate for the ongoing discourse than their competitors. This document is divided into three parts. The first is introductory, and reviews a selection of proposals from the sentence processing literature. The second part explores a body of data which has been adduced in support of a theory of structural preferences --- one that is inconsistent with the present claim. I show how the current proposal can be specified to account for the available data, and moreover to predict where structural preference theories will go wrong. The third part is a theoretical investigation of how well the proposed architecture can be realized using current conceptions of linguistic competence. In it, I present a parsing algorithm and a meaning-based ambiguity resolution method.
cmp-lg/9406030
The complexity of normal form rewrite sequences for Associativity
cmp-lg cs.CL
The complexity of a particular term-rewrite system is considered: the rule of associativity (x*y)*z --> x*(y*z). Algorithms and exact calculations are given for the longest and shortest sequences of applications of --> that result in normal form (NF). The shortest NF sequence for a term x is always n-drm(x), where n is the number of occurrences of * in x and drm(x) is the depth of the rightmost leaf of x. The longest NF sequence for any term is of length n(n-1)/2.
cmp-lg/9406031
A Psycholinguistically Motivated Parser for CCG
cmp-lg cs.CL
Considering the speed in which humans resolve syntactic ambiguity, and the overwhelming evidence that syntactic ambiguity is resolved through selection of the analysis whose interpretation is the most `sensible', one comes to the conclusion that interpretation, hence parsing take place incrementally, just about every word. Considerations of parsimony in the theory of the syntactic processor lead one to explore the simplest of parsers: one which represents only analyses as defined by the grammar and no other information. Toward this aim of a simple, incremental parser I explore the proposal that the competence grammar is a Combinatory Categorial Grammar (CCG). I address the problem of the proliferating analyses that stem from CCG's associativity of derivation. My solution involves maintaining only the maximally incremental analysis and, when necessary, computing the maximally right-branching analysis. I use results from the study of rewrite systems to show that this computation is efficient.
cmp-lg/9406032
Anytime Algorithms for Speech Parsing?
cmp-lg cs.CL
This paper discusses to which extent the concept of ``anytime algorithms'' can be applied to parsing algorithms with feature unification. We first try to give a more precise definition of what an anytime algorithm is. We arque that parsing algorithms have to be classified as contract algorithms as opposed to (truly) interruptible algorithms. With the restriction that the transaction being active at the time an interrupt is issued has to be completed before the interrupt can be executed, it is possible to provide a parser with limited anytime behavior, which is in fact being realized in our research prototype.
cmp-lg/9406033
Verb Semantics and Lexical Selection
cmp-lg cs.CL
This paper will focus on the semantic representation of verbs in computer systems and its impact on lexical selection problems in machine translation (MT). Two groups of English and Chinese verbs are examined to show that lexical selection must be based on interpretation of the sentence as well as selection restrictions placed on the verb arguments. A novel representation scheme is suggested, and is compared to representations with selection restrictions used in transfer-based MT. We see our approach as closely aligned with knowledge-based MT approaches (KBMT), and as a separate component that could be incorporated into existing systems. Examples and experimental results will show that, using this scheme, inexact matches can achieve correct lexical selection.
cmp-lg/9406034
Decision Lists for Lexical Ambiguity Resolution: Application to Accent Restoration in Spanish and French
cmp-lg cs.CL
This paper presents a statistical decision procedure for lexical ambiguity resolution. The algorithm exploits both local syntactic patterns and more distant collocational evidence, generating an efficient, effective, and highly perspicuous recipe for resolving a given ambiguity. By identifying and utilizing only the single best disambiguating evidence in a target context, the algorithm avoids the problematic complex modeling of statistical dependencies. Although directly applicable to a wide class of ambiguities, the algorithm is described and evaluated in a realistic case study, the problem of restoring missing accents in Spanish and French text.
cmp-lg/9406035
DISCO---An HPSG-based NLP System and its Application for Appointment Scheduling (Project Note)
cmp-lg cs.CL
The natural language system DISCO is described. It combines o a powerful and flexible grammar development system; o linguistic competence for German including morphology, syntax and semantics; o new methods for linguistic performance modelling on the basis of high-level competence grammars; o new methods for modelling multi-agent dialogue competence; o an interesting sample application for appointment scheduling and calendar management.
cmp-lg/9406036
Text Analysis Tools in Spoken Language Processing
cmp-lg cs.CL
This submission contains the postscript of the final version of the slides used in our ACL-94 tutorial.
cmp-lg/9406037
Multi-Paragraph Segmentation of Expository Text
cmp-lg cs.CL
This paper describes TextTiling, an algorithm for partitioning expository texts into coherent multi-paragraph discourse units which reflect the subtopic structure of the texts. The algorithm uses domain-independent lexical frequency and distribution information to recognize the interactions of multiple simultaneous themes. Two fully-implemented versions of the algorithm are described and shown to produce segmentation that corresponds well to human judgments of the major subtopic boundaries of thirteen lengthy texts.
cmp-lg/9406038
An Empirical Model of Acknowledgment for Spoken-Language Systems
cmp-lg cs.CL
We refine and extend prior views of the description, purposes, and contexts-of-use of acknowledgment acts through empirical examination of the use of acknowledgments in task-based conversation. We distinguish three broad classes of acknowledgments (other-->ackn, self-->other-->ackn, and self+ackn) and present a catalogue of 13 patterns within these classes that account for the specific uses of acknowledgment in the corpus.
cmp-lg/9406039
Three studies of grammar-based surface-syntactic parsing of unrestricted English text. A summary and orientation
cmp-lg cs.CL
The dissertation addresses the design of parsing grammars for automatic surface-syntactic analysis of unconstrained English text. It consists of a summary and three articles. {\it Morphological disambiguation} documents a grammar for morphological (or part-of-speech) disambiguation of English, done within the Constraint Grammar framework proposed by Fred Karlsson. The disambiguator seeks to discard those of the alternative morphological analyses proposed by the lexical analyser that are contextually illegitimate. The 1,100 constraints express some 23 general, essentially syntactic statements as restrictions on the linear order of morphological tags. The error rate of the morphological disambiguator is about ten times smaller than that of another state-of-the-art probabilistic disambiguator, given that both are allowed to leave some of the hardest ambiguities unresolved. This accuracy suggests the viability of the grammar-based approach to natural language parsing, thus also contributing to the more general debate concerning the viability of probabilistic vs.\ linguistic techniques. {\it Experiments with heuristics} addresses the question of how to resolve those ambiguities that survive the morphological disambiguator. Two approaches are presented and empirically evaluated: (i) heuristic disambiguation constraints and (ii) techniques for learning from the fully disambiguated part of the corpus and then applying this information to resolving remaining ambiguities.
cmp-lg/9406040
Learning unification-based grammars using the Spoken English Corpus
cmp-lg cs.CL
This paper describes a grammar learning system that combines model-based and data-driven learning within a single framework. Our results from learning grammars using the Spoken English Corpus (SEC) suggest that combined model-based and data-driven learning can produce a more plausible grammar than is the case when using either learning style isolation.
cmp-lg/9407001
Morphology with a Null-Interface
cmp-lg cs.CL
We present an integrated architecture for word-level and sentence-level processing in a unification-based paradigm. The core of the system is a CLP implementation of a unification engine for feature structures supporting relational values. In this framework an HPSG-style grammar is implemented. Word-level processing uses X2MorF, a morphological component based on an extended version of two-level morphology. This component is tightly integrated with the grammar as a relation. The advantage of this approach is that morphology and syntax are kept logically autonomous while at the same time minimizing interface problems.
cmp-lg/9407002
Syntactic Analysis by Local Grammars Automata: an Efficient Algorithm
cmp-lg cs.CL
Local grammars can be represented in a very convenient way by automata. This paper describes and illustrates an efficient algorithm for the application of local grammars put in this form to lemmatized texts.
cmp-lg/9407003
Compact Representations by Finite-State Transducers
cmp-lg cs.CL
Finite-state transducers give efficient representations of many Natural Language phenomena. They allow to account for complex lexicon restrictions encountered, without involving the use of a large set of complex rules difficult to analyze. We here show that these representations can be made very compact, indicate how to perform the corresponding minimization, and point out interesting linguistic side-effects of this operation.
cmp-lg/9407004
Japanese word sense disambiguation based on examples of synonyms
cmp-lg cs.CL
(This is not the abstract): The language is Japanese. If your printer does not have fonts for Japases characters, the characters in figures will not be printed out correctly. Dissertation for Bachelor's degree at Kyoto University(Nagao lab.),March 1994.
cmp-lg/9407005
A Corrective Training Algorithm for Adaptive Learning in Bag Generation
cmp-lg cs.CL
The sampling problem in training corpus is one of the major sources of errors in corpus-based applications. This paper proposes a corrective training algorithm to best-fit the run-time context domain in the application of bag generation. It shows which objects to be adjusted and how to adjust their probabilities. The resulting techniques are greatly simplified and the experimental results demonstrate the promising effects of the training algorithm from generic domain to specific domain. In general, these techniques can be easily extended to various language models and corpus-based applications.
cmp-lg/9407006
Interleaving Syntax and Semantics in an Efficient Bottom-Up Parser
cmp-lg cs.CL
We describe an efficient bottom-up parser that interleaves syntactic and semantic structure building. Two techniques are presented for reducing search by reducing local ambiguity: Limited left-context constraints are used to reduce local syntactic ambiguity, and deferred sortal-constraint application is used to reduce local semantic ambiguity. We experimentally evaluate these techniques, and show dramatic reductions in both number of chart-edges and total parsing time. The robust processing capabilities of the parser are demonstrated in its use in improving the accuracy of a speech recognizer.
cmp-lg/9407007
GEMINI: A Natural Language System for Spoken-Language Understanding
cmp-lg cs.CL
Gemini is a natural language understanding system developed for spoken language applications. The paper describes the architecture of Gemini, paying particular attention to resolving the tension between robustness and overgeneration. Gemini features a broad-coverage unification-based grammar of English, fully interleaved syntactic and semantic processing in an all-paths, bottom-up parser, and an utterance-level parser to find interpretations of sentences that might not be analyzable as complete sentences. Gemini also includes novel components for recognizing and correcting grammatical disfluencies, and for doing parse preferences. This paper presents a component-by-component view of Gemini, providing detailed relevant measurements of size, efficiency, and performance.
cmp-lg/9407008
Tricolor DAGs for Machine Translation
cmp-lg cs.CL
Machine translation (MT) has recently been formulated in terms of constraint-based knowledge representation and unification theories, but it is becoming more and more evident that it is not possible to design a practical MT system without an adequate method of handling mismatches between semantic representations in the source and target languages. In this paper, we introduce the idea of ``information-based'' MT, which is considerably more flexible than interlingual MT or the conventional transfer-based MT.
cmp-lg/9407009
Estimating Performance of Pipelined Spoken Language Translation Systems
cmp-lg cs.CL
Most spoken language translation systems developed to date rely on a pipelined architecture, in which the main stages are speech recognition, linguistic analysis, transfer, generation and speech synthesis. When making projections of error rates for systems of this kind, it is natural to assume that the error rates for the individual components are independent, making the system accuracy the product of the component accuracies. The paper reports experiments carried out using the SRI-SICS-Telia Research Spoken Language Translator and a 1000-utterance sample of unseen data. The results suggest that the naive performance model leads to serious overestimates of system error rates, since there are in fact strong dependencies between the components. Predicting the system error rate on the independence assumption by simple multiplication resulted in a 16\% proportional overestimate for all utterances, and a 19\% overestimate when only utterances of length 1-10 words were considered.
cmp-lg/9407010
Combining Knowledge Sources to Reorder N-Best Speech Hypothesis Lists
cmp-lg cs.CL
A simple and general method is described that can combine different knowledge sources to reorder N-best lists of hypotheses produced by a speech recognizer. The method is automatically trainable, acquiring information from both positive and negative examples. Experiments are described in which it was tested on a 1000-utterance sample of unseen ATIS data.
cmp-lg/9407011
Discourse Obligations in Dialogue Processing
cmp-lg cs.CL
We show that in modeling social interaction, particularly dialogue, the attitude of obligation can be a useful adjunct to the popularly considered attitudes of belief, goal, and intention and their mutual and shared counterparts. In particular, we show how discourse obligations can be used to account in a natural manner for the connection between a question and its answer in dialogue and how obligations can be used along with other parts of the discourse context to extend the coverage of a dialogue system.
cmp-lg/9407012
Phoneme Recognition Using Acoustic Events
cmp-lg cs.CL
This paper presents a new approach to phoneme recognition using nonsequential sub--phoneme units. These units are called acoustic events and are phonologically meaningful as well as recognizable from speech signals. Acoustic events form a phonologically incomplete representation as compared to distinctive features. This problem may partly be overcome by incorporating phonological constraints. Currently, 24 binary events describing manner and place of articulation, vowel quality and voicing are used to recognize all German phonemes. Phoneme recognition in this paradigm consists of two steps: After the acoustic events have been determined from the speech signal, a phonological parser is used to generate syllable and phoneme hypotheses from the event lattice. Results obtained on a speaker--dependent corpus are presented.
cmp-lg/9407013
The Acquisition of a Lexicon from Paired Phoneme Sequences and Semantic Representations
cmp-lg cs.CL
We present an algorithm that acquires words (pairings of phonological forms and semantic representations) from larger utterances of unsegmented phoneme sequences and semantic representations. The algorithm maintains from utterance to utterance only a single coherent dictionary, and learns in the presence of homonymy, synonymy, and noise. Test results over a corpus of utterances generated from the Childes database of mother-child interactions are presented.
cmp-lg/9407014
Abstract Machine for Typed Feature Structures
cmp-lg cs.CL
This paper describes a first step towards the definition of an abstract machine for linguistic formalisms that are based on typed feature structures, such as HPSG. The core design of the abstract machine is given in detail, including the compilation process from a high-level specification language to the abstract machine language and the implementation of the abstract instructions. We thus apply methods that were proved useful in computer science to the study of natural languages: a grammar specified using the formalism is endowed with an operational semantics. Currently, our machine supports the unification of simple feature structures, unification of sequences of such structures, cyclic structures and disjunction.
cmp-lg/9407015
Specifying Intonation from Context for Speech Synthesis
cmp-lg cs.CL
This paper presents a theory and a computational implementation for generating prosodically appropriate synthetic speech in response to database queries. Proper distinctions of contrast and emphasis are expressed in an intonation contour that is synthesized by rule under the control of a grammar, a discourse model, and a knowledge base. The theory is based on Combinatory Categorial Grammar, a formalism which easily integrates the notions of syntactic constituency, semantics, prosodic phrasing and information structure. Results from our current implementation demonstrate the system's ability to generate a variety of intonational possibilities for a given sentence depending on the discourse context.
cmp-lg/9407016
The Role of Cognitive Modeling in Achieving Communicative Intentions
cmp-lg cs.CL
A discourse planner for (task-oriented) dialogue must be able to make choices about whether relevant, but optional information (for example, the "satellites" in an RST-based planner) should be communicated. We claim that effective text planners must explicitly model aspects of the Hearer's cognitive state, such as what the hearer is attending to and what inferences the hearer can draw, in order to make these choices. We argue that a mere representation of the Hearer's knowledge is inadequate. We support this claim by (1) an analysis of naturally occurring dialogue, and (2) by simulating the generation of discourses in a situation in which we can vary the cognitive parameters of the hearer. Our results show that modeling cognitive state can lead to more effective discourses (measured with respect to a simple task).
cmp-lg/9407017
Generating Context-Appropriate Word Orders in Turkish
cmp-lg cs.CL
Turkish has considerably freer word order than English. The interpretations of different word orders in Turkish rely on information that describes how a sentence relates to its discourse context. To capture the syntactic features of a free word order language, I present an adaptation of Combinatory Categorial Grammars called {}-CCGs (set-CCGs). In {}-CCGs, a verb's subcategorization requirements are relaxed so that it requires a set of arguments without specifying their linear order. I integrate a level of information structure, representing pragmatic functions such as topic and focus, with {}-CCGs to allow certain pragmatic distinctions in meaning to influence the word order of a sentence in a compositional way. Finally, I discuss how this strategy is used within an implemented generation system which produces Turkish sentences with context-appropriate word orders in a simple database query task.
cmp-lg/9407018
Generating Multilingual Documents from a Knowledge Base: The TECHDOC Project
cmp-lg cs.CL
TECHDOC is an implemented system demonstrating the feasibility of generating multilingual technical documents on the basis of a language-independent knowledge base. Its application domain is user and maintenance instructions, which are produced from underlying plan structures representing the activities, the participating objects with their properties, relations, and so on. This paper gives a brief outline of the system architecture and discusses some recent developments in the project: the addition of actual event simulation in the KB, steps towards a document authoring tool, and a multimodal user interface. (slightly corrected version of a paper to appear in: COLING 94, Proceedings)
cmp-lg/9407019
Tracking Point of View in Narrative
cmp-lg cs.CL
Third-person fictional narrative text is composed not only of passages that objectively narrate events, but also of passages that present characters' thoughts, perceptions, and inner states. Such passages take a character's ``psychological point of view''. A language understander must determine the current psychological point of view in order to distinguish the beliefs of the characters from the facts of the story, to correctly attribute beliefs and other attitudes to their sources, and to understand the discourse relations among sentences. Tracking the psychological point of view is not a trivial problem, because many sentences are not explicitly marked for point of view, and whether the point of view of a sentence is objective or that of a character (and if the latter, which character it is) often depends on the context in which the sentence appears. Tracking the psychological point of view is the problem addressed in this work. The approach is to seek, by extensive examinations of naturally-occurring narrative, regularities in the ways that authors manipulate point of view, and to develop an algorithm that tracks point of view on the basis of the regularities found. This paper presents this algorithm, gives demonstrations of an implemented system, and describes the results of some preliminary empirical studies, which lend support to the algorithm.
cmp-lg/9407020
A Sequential Algorithm for Training Text Classifiers
cmp-lg cs.CL
The ability to cheaply train text classifiers is critical to their use in information retrieval, content analysis, natural language processing, and other tasks involving data which is partly or fully textual. An algorithm for sequential sampling during machine learning of statistical classifiers was developed and tested on a newswire text categorization task. This method, which we call uncertainty sampling, reduced by as much as 500-fold the amount of training data that would have to be manually classified to achieve a given level of effectiveness.
cmp-lg/9407021
K-vec: A New Approach for Aligning Parallel Texts
cmp-lg cs.CL
Various methods have been proposed for aligning texts in two or more languages such as the Canadian Parliamentary Debates(Hansards). Some of these methods generate a bilingual lexicon as a by-product. We present an alternative alignment strategy which we call K-vec, that starts by estimating the lexicon. For example, it discovers that the English word "fisheries" is similar to the French "pe^ches" by noting that the distribution of "fisheries" in the English text is similar to the distribution of "pe^ches" in the French. K-vec does not depend on sentence boundaries.
cmp-lg/9407022
Comparative Discourse Analysis of Parallel Texts
cmp-lg cs.CL
A quantitative representation of discourse structure can be computed by measuring lexical cohesion relations among adjacent blocks of text. These representations have been proposed to deal with sub-topic text segmentation. In a parallel corpus, similar representations can be derived for versions of a text in various languages. These can be used for parallel segmentation and as an alternative measure of text-translation similarity.
cmp-lg/9407023
Multi-Tape Two-Level Morphology: A Case Study in Semitic Non-linear Morphology
cmp-lg cs.CL
This paper presents an implemented multi-tape two-level model capable of describing Semitic non-linear morphology. The computational framework behind the current work is motivated by Kay (1987); the formalism presented here is an extension to the formalism reported by Pulman and Hepple (1993). The objectives of the current work are: to stay as close as possible, in spirit, to standard two-level morphology, to stay close to the linguistic description of Semitic stems, and to present a model which can be used with ease by the Semitist. The paper illustrates that if finite-state transducers (FSTs) in a standard two-level morphology model are replaced with multi-tape auxiliary versions (AFSTs), one can account for Semitic root-and-pattern morphology using high level notation.
cmp-lg/9407024
PRINCIPAR---An Efficient, Broad-coverage, Principle-based Parser
cmp-lg cs.CL
We present an efficient, broad-coverage, principle-based parser for English. The parser has been implemented in C++ and runs on SUN Sparcstations with X-windows. It contains a lexicon with over 90,000 entries, constructed automatically by applying a set of extraction and conversion rules to entries from machine readable dictionaries.
cmp-lg/9407025
Recovering From Parser Failures: A Hybrid Statistical/Symbolic Approach
cmp-lg cs.CL
We describe an implementation of a hybrid statistical/symbolic approach to repairing parser failures in a speech-to-speech translation system. We describe a module which takes as input a fragmented parse and returns a repaired meaning representation. It negotiates with the speaker about what the complete meaning of the utterance is by generating hypotheses about how to fit the fragments of the partial parse together into a coherent meaning representation. By drawing upon both statistical and symbolic information, it constrains its repair hypotheses to those which are both likely and meaningful. Because it updates its statistical model during use, it improves its performance over time.
cmp-lg/9407026
Tagging and Morphological Disambiguation of Turkish Text
cmp-lg cs.CL
Automatic text tagging is an important component in higher level analysis of text corpora, and its output can be used in many natural language processing applications. In languages like Turkish or Finnish, with agglutinative morphology, morphological disambiguation is a very crucial process in tagging, as the structures of many lexical forms are morphologically ambiguous. This paper describes a POS tagger for Turkish text based on a full-scale two-level specification of Turkish morphology that is based on a lexicon of about 24,000 root words. This is augmented with a multi-word and idiomatic construct recognizer, and most importantly morphological disambiguator based on local neighborhood constraints, heuristics and limited amount of statistical information. The tagger also has functionality for statistics compilation and fine tuning of the morphological analyzer, such as logging erroneous morphological parses, commonly used roots, etc. Preliminary results indicate that the tagger can tag about 98-99\% of the texts accurately with very minimal user intervention. Furthermore for sentences morphologically disambiguated with the tagger, an LFG parser developed for Turkish, generates, on the average, 50\% less ambiguous parses and parses almost 2.5 times faster. The tagging functionality is not specific to Turkish, and can be applied to any language with a proper morphological analysis interface.
cmp-lg/9407027
Parsing as Tree Traversal
cmp-lg cs.CL
This paper presents a unified approach to parsing, in which top-down, bottom-up and left-corner parsers are related to preorder, postorder and inorder tree traversals. It is shown that the simplest bottom-up and left-corner parsers are left recursive and must be converted using an extended Greibach normal form. With further partial execution, the bottom-up and left-corner parsers collapse together as in the BUP parser of Matsumoto.
cmp-lg/9407028
Automated Postediting of Documents
cmp-lg cs.CL
Large amounts of low- to medium-quality English texts are now being produced by machine translation (MT) systems, optical character readers (OCR), and non-native speakers of English. Most of this text must be postedited by hand before it sees the light of day. Improving text quality is tedious work, but its automation has not received much research attention. Anyone who has postedited a technical report or thesis written by a non-native speaker of English knows the potential of an automated postediting system. For the case of MT-generated text, we argue for the construction of postediting modules that are portable across MT systems, as an alternative to hardcoding improvements inside any one system. As an example, we have built a complete self-contained postediting module for the task of article selection (a, an, the) for English noun phrases. This is a notoriously difficult problem for Japanese-English MT. Our system contains over 200,000 rules derived automatically from online text resources. We report on learning algorithms, accuracy, and comparisons with human performance.
cmp-lg/9407029
Building a Large-Scale Knowledge Base for Machine Translation
cmp-lg cs.CL
Knowledge-based machine translation (KBMT) systems have achieved excellent results in constrained domains, but have not yet scaled up to newspaper text. The reason is that knowledge resources (lexicons, grammar rules, world models) must be painstakingly handcrafted from scratch. One of the hypotheses being tested in the PANGLOSS machine translation project is whether or not these resources can be semi-automatically acquired on a very large scale. This paper focuses on the construction of a large ontology (or knowledge base, or world model) for supporting KBMT. It contains representations for some 70,000 commonly encountered objects, processes, qualities, and relations. The ontology was constructed by merging various online dictionaries, semantic networks, and bilingual resources, through semi-automatic methods. Some of these methods (e.g., conceptual matching of semantic taxonomies) are broadly applicable to problems of importing/exporting knowledge from one KB to another. Other methods (e.g., bilingual matching) allow a knowledge engineer to build up an index to a KB in a second language, such as Spanish or Japanese.
cmp-lg/9407030
Computing FIRST and FOLLOW Functions for Feature-Theoretic Grammars
cmp-lg cs.CL
This paper describes an algorithm for the computation of FIRST and FOLLOW sets for use with feature-theoretic grammars in which the value of the sets consists of pairs of feature-theoretic categories. The algorithm preserves as much information from the grammars as possible, using negative restriction to define equivalence classes. Addition of a simple data structure leads to an order of magnitude improvement in execution time over a naive implementation.
cmp-lg/9408001
The Correct and Efficient Implementation of Appropriateness Specifications for Typed Feature Structures
cmp-lg cs.CL
In this paper, we argue that type inferencing incorrectly implements appropriateness specifications for typed feature structures, promote a combination of type resolution and unfilling as a correct and efficient alternative, and consider the expressive limits of this alternative approach. Throughout, we use feature cooccurence restrictions as illustration and linguistic motivation.
cmp-lg/9408002
Computational Analyses of Arabic Morphology
cmp-lg cs.CL
This paper demonstrates how a (multi-tape) two-level formalism can be used to write two-level grammars for Arabic non-linear morphology using a high level, but computationally tractable, notation. Three illustrative grammars are provided based on CV-, moraic- and affixational analyses. These are complemented by a proposal for handling the hitherto computationally untreated problem of the broken plural. It will be shown that the best grammars for describing Arabic non-linear morphology are moraic in the case of templatic stems, and affixational in the case of a-templatic stems. The paper will demonstrate how the broken plural can be derived under two-level theory via the `implicit' derivation of the singular.
cmp-lg/9408003
Typed Feature Structures as Descriptions
cmp-lg cs.CL
A description is an entity that can be interpreted as true or false of an object, and using feature structures as descriptions accrues several computational benefits. In this paper, I create an explicit interpretation of a typed feature structure used as a description, define the notion of a satisfiable feature structure, and create a simple and effective algorithm to decide if a feature structure is satisfiable.
cmp-lg/9408004
Parsing with Principles and Probabilities
cmp-lg cs.CL
This paper is an attempt to bring together two approaches to language analysis. The possible use of probabilistic information in principle-based grammars and parsers is considered, including discussion on some theoretical and computational problems that arise. Finally a partial implementation of these ideas is presented, along with some preliminary results from testing on a small set of sentences.
cmp-lg/9408005
A Modular and Flexible Architecture for an Integrated Corpus Query System
cmp-lg cs.CL
The paper describes the architecture of an integrated and extensible corpus query system developed at the University of Stuttgart and gives examples of some of the modules realized within this architecture. The modules form the core of a corpus workbench. Within the proposed architecture, information required for the evaluation of queries may be derived from different knowledge sources (the corpus text, databases, on-line thesauri) and by different means: either through direct lookup in a database or by calling external tools which may infer the necessary information at the time of query evaluation. The information available and the method of information access can be stated declaratively and individually for each corpus, leading to a flexible, extensible and modular corpus workbench.
cmp-lg/9408006
LHIP: Extended DCGs for Configurable Robust Parsing
cmp-lg cs.CL
We present LHIP, a system for incremental grammar development using an extended DCG formalism. The system uses a robust island-based parsing method controlled by user-defined performance thresholds.
cmp-lg/9408007
Emergent Linguistic Rules from Inducing Decision Trees: Disambiguating Discourse Clue Words
cmp-lg cs.CL
We apply decision tree induction to the problem of discourse clue word sense disambiguation with a genetic algorithm. The automatic partitioning of the training set which is intrinsic to decision tree induction gives rise to linguistically viable rules.
cmp-lg/9408008
Statistical versus symbolic parsing for captioned-information retrieval
cmp-lg cs.CL
We discuss implementation issues of MARIE-1, a mostly symbolic parser fully implemented, and MARIE-2, a more statistical parser partially implemented. They address a corpus of 100,000 picture captions. We argue that the mixed approach of MARIE-2 should be better for this corpus because its algorithms (not data) are simpler.
cmp-lg/9408009
Tagging accurately -- Don't guess if you know
cmp-lg cs.CL
We discuss combining knowledge-based (or rule-based) and statistical part-of-speech taggers. We use two mature taggers, ENGCG and Xerox Tagger, to independently tag the same text and combine the results to produce a fully disambiguated text. In a 27000 word test sample taken from a previously unseen corpus we achieve 98.5% accuracy. This paper presents the data in detail. We describe the problems we encountered in the course of combining the two taggers and discuss the problem of evaluating taggers.
cmp-lg/9408010
On Using Selectional Restriction in Language Models for Speech Recognition
cmp-lg cs.CL
In this paper, we investigate the use of selectional restriction -- the constraints a predicate imposes on its arguments -- in a language model for speech recognition. We use an un-tagged corpus, followed by a public domain tagger and a very simple finite state machine to obtain verb-object pairs from unrestricted English text. We then measure the impact the knowledge of the verb has on the prediction of the direct object in terms of the perplexity of a cluster-based language model. The results show that even though a clustered bigram is more useful than a verb-object model, the combination of the two leads to an improvement over the clustered bigram model.
cmp-lg/9408011
Distributional Clustering of English Words
cmp-lg cs.CL
We describe and experimentally evaluate a method for automatically clustering words according to their distribution in particular syntactic contexts. Deterministic annealing is used to find lowest distortion sets of clusters. As the annealing parameter increases, existing clusters become unstable and subdivide, yielding a hierarchical ``soft'' clustering of the data. Clusters are used as the basis for class models of word coocurrence, and the models evaluated with respect to held-out test data.
cmp-lg/9408012
Approximate N-Gram Markov Model for Natural Language Generation
cmp-lg cs.CL
This paper proposes an Approximate n-gram Markov Model for bag generation. Directed word association pairs with distances are used to approximate (n-1)-gram and n-gram training tables. This model has parameters of word association model, and merits of both word association model and Markov Model. The training knowledge for bag generation can be also applied to lexical selection in machine translation design.
cmp-lg/9408013
Training and Scaling Preference Functions for Disambiguation
cmp-lg cs.CL
We present an automatic method for weighting the contributions of preference functions used in disambiguation. Initial scaling factors are derived as the solution to a least-squares minimization problem, and improvements are then made by hill-climbing. The method is applied to disambiguating sentences in the ATIS (Air Travel Information System) corpus, and the performance of the resulting scaling factors is compared with hand-tuned factors. We then focus on one class of preference function, those based on semantic lexical collocations. Experimental results are presented showing that such functions vary considerably in selecting correct analyses. In particular we define a function that performs significantly better than ones based on mutual information and likelihood ratios of lexical associations.
cmp-lg/9408014
Qualitative and Quantitative Models of Speech Translation
cmp-lg cs.CL
This paper compares a qualitative reasoning model of translation with a quantitative statistical model. We consider these models within the context of two hypothetical speech translation systems, starting with a logic-based design and pointing out which of its characteristics are best preserved or eliminated in moving to the second, quantitative design. The quantitative language and translation models are based on relations between lexical heads of phrases. Statistical parameters for structural dependency, lexical transfer, and linear order are used to select a set of implicit relations between words in a source utterance, a corresponding set of relations between target language words, and the most likely translation of the original utterance.
cmp-lg/9408015
Experimentally Evaluating Communicative Strategies: The Effect of the Task
cmp-lg cs.CL
Effective problem solving among multiple agents requires a better understanding of the role of communication in collaboration. In this paper we show that there are communicative strategies that greatly improve the performance of resource-bounded agents, but that these strategies are highly sensitive to the task requirements, situation parameters and agents' resource limitations. We base our argument on two sources of evidence: (1) an analysis of a corpus of 55 problem solving dialogues, and (2) experimental simulations of collaborative problem solving dialogues in an experimental world, Design-World, where we parameterize task requirements, agents' resources and communicative strategies.
cmp-lg/9408016
On Implementing an HPSG theory -- Aspects of the logical architecture, the formalization, and the implementation of head-driven phrase structure grammars
cmp-lg cs.CL
The paper presents some aspects involved in the formalization and implementation of HPSG theories. As basis, the logical setups of Carpenter (1992) and King (1989, 1994) are briefly compared regarding their usefulness as basis for HPSGII (Pollard and Sag 1994). The possibilities for expressing HPSG theories in the HPSGII architecture and in various computational systems (ALE, Troll, CUF, and TFS) are discussed. Beside a formal characterization of the possibilities, the paper investigates the specific choices for constraints with certain linguistic motivations, i.e. the lexicon, structure licencing, and grammatical principles. An ALE implementation of a theory for German proposed by Hinrichs and Nakazawa (1994) is used as example and the ALE grammar is included in the appendix.
cmp-lg/9408017
Reaping the Benefits of Interactive Syntax and Semantics
cmp-lg cs.CL
Semantic feedback is an important source of information that a parser could use to deal with local ambiguities in syntax. However, it is difficult to devise a systematic communication mechanism for interactive syntax and semantics. In this article, I propose a variant of left-corner parsing to define the points at which syntax and semantics should interact, an account of grammatical relations and thematic roles to define the content of the communication, and a conflict resolution strategy based on independent preferences from syntax and semantics. The resulting interactive model has been implemented in a program called COMPERE and shown to account for a wide variety of psycholinguistic data on structural and lexical ambiguities.
cmp-lg/9408018
Uniform Representations for Syntax-Semantics Arbitration
cmp-lg cs.CL
Psychological investigations have led to considerable insight into the working of the human language comprehension system. In this article, we look at a set of principles derived from psychological findings to argue for a particular organization of linguistic knowledge along with a particular processing strategy and present a computational model of sentence processing based on those principles. Many studies have shown that human sentence comprehension is an incremental and interactive process in which semantic and other higher-level information interacts with syntactic information to make informed commitments as early as possible at a local ambiguity. Early commitments may be made by using top-down guidance from knowledge of different types, each of which must be applicable independently of others. Further evidence from studies of error recovery and delayed decisions points toward an arbitration mechanism for combining syntactic and semantic information in resolving ambiguities. In order to account for all of the above, we propose that all types of linguistic knowledge must be represented in a common form but must be separable so that they can be applied independently of each other and integrated at processing time by the arbitrator. We present such a uniform representation and a computational model called COMPERE based on the representation and the processing strategy.
cmp-lg/9408019
Building a Parser That can Afford to Interact with Semantics
cmp-lg cs.CL
Natural language understanding programs get bogged down by the multiplicity of possible syntactic structures while processing real world texts that human understanders do not have much difficulty with. In this work, I analyze the relationships between parsing strategies, the degree of local ambiguity encountered by them, and semantic feedback to syntax, and propose a parsing algorithm called {\em Head-Signaled Left Corner Parsing} (HSLC) that minimizes local ambiguities while supporting interactive syntactic and semantic analysis. Such a parser has been implemented in a sentence understanding program called COMPERE.
cmp-lg/9408020
Having Your Cake and Eating It Too: Autonomy and Interaction in a Model of Sentence Processing
cmp-lg cs.CL
Is the human language understander a collection of modular processes operating with relative autonomy, or is it a single integrated process? This ongoing debate has polarized the language processing community, with two fundamentally different types of model posited, and with each camp concluding that the other is wrong. One camp puts forth a model with separate processors and distinct knowledge sources to explain one body of data, and the other proposes a model with a single processor and a homogeneous, monolithic knowledge source to explain the other body of data. In this paper we argue that a hybrid approach which combines a unified processor with separate knowledge sources provides an explanation of both bodies of data, and we demonstrate the feasibility of this approach with the computational model called COMPERE. We believe that this approach brings the language processing community significantly closer to offering human-like language processing systems.
cmp-lg/9408021
A Unified Process Model of Syntactic and Semantic Error Recovery in Sentence Understanding
cmp-lg cs.CL
The development of models of human sentence processing has traditionally followed one of two paths. Either the model posited a sequence of processing modules, each with its own task-specific knowledge (e.g., syntax and semantics), or it posited a single processor utilizing different types of knowledge inextricably integrated into a monolithic knowledge base. Our previous work in modeling the sentence processor resulted in a model in which different processing modules used separate knowledge sources but operated in parallel to arrive at the interpretation of a sentence. One highlight of this model is that it offered an explanation of how the sentence processor might recover from an error in choosing the meaning of an ambiguous word. Recent experimental work by Laurie Stowe strongly suggests that the human sentence processor deals with syntactic error recovery using a mechanism very much like that proposed by our model of semantic error recovery. Another way to interpret Stowe's finding is this: the human sentence processor consists of a single unified processing module utilizing multiple independent knowledge sources in parallel. A sentence processor built upon this architecture should at times exhibit behavior associated with modular approaches, and at other times act like an integrated system. In this paper we explore some of these ideas via a prototype computational model of sentence processing called COMPERE, and propose a set of psychological experiments for testing our theories.
cmp-lg/9409001
Integrating Knowledge Bases and Statistics in MT
cmp-lg cs.CL
We summarize recent machine translation (MT) research at the Information Sciences Institute of USC, and we describe its application to the development of a Japanese-English newspaper MT system. Our work aims at scaling up grammar-based, knowledge-based MT techniques. This scale-up involves the use of statistical methods, both in acquiring effective knowledge resources and in making reasonable linguistic choices in the face of knowledge gaps.
cmp-lg/9409002
Conceptual Association for Compound Noun Analysis
cmp-lg cs.CL
This paper describes research toward the automatic interpretation of compound nouns using corpus statistics. An initial study aimed at syntactic disambiguation is presented. The approach presented bases associations upon thesaurus categories. Association data is gathered from unambiguous cases extracted from a corpus and is then applied to the analysis of ambiguous compound nouns. While the work presented is still in progress, a first attempt to syntactically analyse a test set of 244 examples shows 75% correctness. Future work is aimed at improving this accuracy and extending the technique to assign semantic role information, thus producing a complete interpretation.
cmp-lg/9409003
A Probabilistic Model of Compound Nouns
cmp-lg cs.CL
Compound nouns such as example noun compound are becoming more common in natural language and pose a number of difficult problems for NLP systems, notably increasing the complexity of parsing. In this paper we develop a probabilistic model for syntactically analysing such compounds. The model predicts compound noun structures based on knowledge of affinities between nouns, which can be acquired from a corpus. Problems inherent in this corpus-based approach are addressed: data sparseness is overcome by the use of semantically motivated word classes and sense ambiguity is explicitly handled in the model. An implementation based on this model is described in Lauer (1994) and correctly parses 77% of the test set.
cmp-lg/9409004
An Experiment on Learning Appropriate Selectional Restrictions from a Parsed Corpus
cmp-lg cs.CL
We present a methodology to extract Selectional Restrictions at a variable level of abstraction from phrasally analyzed corpora. The method relays in the use of a wide-coverage noun taxonomy and a statistical measure of the co-occurrence of linguistic items. Some experimental results about the performance of the method are provided.
cmp-lg/9409005
Focusing for Pronoun Resolution in English Discourse: An Implementation
cmp-lg cs.CL
Anaphora resolution is one of the most active research areas in natural language processing. This study examines focusing as a tool for the resolution of pronouns which are a kind of anaphora. Focusing is a discourse phenomenon like anaphora. Candy Sidner formalized focusing in her 1979 MIT PhD thesis and devised several algorithms to resolve definite anaphora including pronouns. She presented her theory in a computational framework but did not generally implement the algorithms. Her algorithms related to focusing and pronoun resolution are implemented in this thesis. This implementation provides a better comprehension of the theory both from a conceptual and a computational point of view. The resulting program is tested on different discourse segments, and evaluation and analysis of the experiments are presented together with the statistical results.
cmp-lg/9409006
Situated Modeling of Epistemic Puzzles
cmp-lg cs.CL
Situation theory is a mathematical theory of meaning introduced by Jon Barwise and John Perry. It has evoked great theoretical and practical interest and motivated the framework of a few `computational' systems. PROSIT is the pioneering work in this direction. Unfortunately, there is a lack of real-life applications on these systems and this study is a preliminary attempt to remedy this deficiency. Here, we examine how much PROSIT reflects situation-theoretic concepts and solve a group of epistemic puzzles using the constructs provided by this programming language.
cmp-lg/9409007
Treating `Free Word Order' in Machine Translation
cmp-lg cs.CL
In `free word order' languages, every sentence is embedded in its specific context. Among others, the order of constituents is determined by the categories `theme', `rheme' and `contrastive focus'. This paper shows how to recognise and to translate these categories automatically on a sentential basis, so that sentence embedding can be achieved without having to refer to the context. Modifier classes, which are traditionally neglected in linguistic description, are fully covered by the proposed method. (Coling 94, Kyoto, Vol. I, pages 69-75)
cmp-lg/9409008
Parsing of Spoken Language under Time Constraints
cmp-lg cs.CL
Spoken language applications in natural dialogue settings place serious requirements on the choice of processing architecture. Especially under adverse phonetic and acoustic conditions parsing procedures have to be developed which do not only analyse the incoming speech in a time-synchroneous and incremental manner, but which are able to schedule their resources according to the varying conditions of the recognition process. Depending on the actual degree of local ambiguity the parser has to select among the available constraints in order to narrow down the search space with as little effort as possible. A parsing approach based on constraint satisfaction techniques is discussed. It provides important characteristics of the desired real-time behaviour and attempts to mimic some of the attention focussing capabilities of the human speech comprehension mechanism.
cmp-lg/9409009
Linguistics Computation, Automatic Model Generation, and Intensions
cmp-lg cs.CL
Techniques are presented for defining models of computational linguistics theories. The methods of generalized diagrams that were developed by this author for modeling artificial intelligence planning and reasoning are shown to be applicable to models of computation of linguistics theories. It is shown that for extensional and intensional interpretations, models can be generated automatically which assign meaning to computations of linguistics theories for natural languages. Keywords: Computational Linguistics, Reasoning Models, G-diagrams For Models, Dynamic Model Implementation, Linguistics and Logics For Artificial Intelligence
cmp-lg/9409010
Inducing Probabilistic Grammars by Bayesian Model Merging
cmp-lg cs.CL
We describe a framework for inducing probabilistic grammars from corpora of positive samples. First, samples are {\em incorporated} by adding ad-hoc rules to a working grammar; subsequently, elements of the model (such as states or nonterminals) are {\em merged} to achieve generalization and a more compact representation. The choice of what to merge and when to stop is governed by the Bayesian posterior probability of the grammar given the data, which formalizes a trade-off between a close fit to the data and a default preference for simpler models (`Occam's Razor'). The general scheme is illustrated using three types of probabilistic grammars: Hidden Markov models, class-based $n$-grams, and stochastic context-free grammars.
cmp-lg/9409011
Aligning Noisy Parallel Corpora Across Language Groups : Word Pair Feature Matching by Dynamic Time Warping
cmp-lg cs.CL
We propose a new algorithm called DK-vec for aligning pairs of Asian/Indo-European noisy parallel texts without sentence boundaries. DK-vec improves on previous alignment algorithms in that it handles better the non-linear nature of noisy corpora. The algorithm uses frequency, position and recency information as features for pattern matching. Dynamic Time Warping is used as the matching technique between word pairs. This algorithm produces a small bilingual lexicon which provides anchor points for alignment.
cmp-lg/9409012
Towards an Automatic Dictation System for Translators: the TransTalk Project
cmp-lg cs.CL
Professional translators often dictate their translations orally and have them typed afterwards. The TransTalk project aims at automating the second part of this process. Its originality as a dictation system lies in the fact that both the acoustic signal produced by the translator and the source text under translation are made available to the system. Probable translations of the source text can be predicted and these predictions used to help the speech recognition system in its lexical choices. We present the results of the first prototype, which show a marked improvement in the performance of the speech recognition task when translation predictions are taken into account.
cmp-lg/9410001
Improving Language Models by Clustering Training Sentences
cmp-lg cs.CL
Many of the kinds of language model used in speech understanding suffer from imperfect modeling of intra-sentential contextual influences. I argue that this problem can be addressed by clustering the sentences in a training corpus automatically into subcorpora on the criterion of entropy reduction, and calculating separate language model parameters for each cluster. This kind of clustering offers a way to represent important contextual effects and can therefore significantly improve the performance of a model. It also offers a reasonably automatic means to gather evidence on whether a more complex, context-sensitive model using the same general kind of linguistic information is likely to reward the effort that would be required to develop it: if clustering improves the performance of a model, this proves the existence of further context dependencies, not exploited by the unclustered model. As evidence for these claims, I present results showing that clustering improves some models but not others for the ATIS domain. These results are consistent with other findings for such models, suggesting that the existence or otherwise of an improvement brought about by clustering is indeed a good pointer to whether it is worth developing further the unclustered model.
cmp-lg/9410002
Lexikoneintraege fuer deutsche Adverbien (Dictionary Entries for German Adverbs)
cmp-lg cs.CL
Modifiers in general, and adverbs in particular, are neglected categories in linguistics, and consequently, their treatment in Natural Language Processing poses problems. In this article, we present the dictionary information for German adverbs which is necessary to deal with word order, degree modifier scope and other problems in NLP. We also give evidence for the claim that a classification according to position classes differs from any semantic classification.
cmp-lg/9410003
Principle Based Semantics for HPSG
cmp-lg cs.CL
The paper presents a constraint based semantic formalism for HPSG. The advantages of the formlism are shown with respect to a grammar for a fragment of German that deals with (i) quantifier scope ambiguities triggered by scrambling and/or movement and (ii) ambiguities that arise from the collective/distributive distinction of plural NPs. The syntax-semantics interface directly implements syntactic conditions on quantifier scoping and distributivity. The construction of semantic representations is guided by general principles governing the interaction between syntax and semantics. Each of these principles acts as a constraint to narrow down the set of possible interpretations of a sentence. Meanings of ambiguous sentences are represented by single partial representations (so-called U(nderspecified) D(iscourse) R(epresentation) S(tructure)s) to which further constraints can be added monotonically to gain more information about the content of a sentence. There is no need to build up a large number of alternative representations of the sentence which are then filtered by subsequent discourse and world knowledge. The advantage of UDRSs is not only that they allow for monotonic incremental interpretation but also that they are equipped with truth conditions and a proof theory that allows for inferences to be drawn directly on structures where quantifier scope is not resolved.
cmp-lg/9410004
Spelling Correction in Agglutinative Languages
cmp-lg cs.CL
This paper presents an approach to spelling correction in agglutinative languages that is based on two-level morphology and a dynamic programming based search algorithm. Spelling correction in agglutinative languages is significantly different than in languages like English. The concept of a word in such languages is much wider that the entries found in a dictionary, owing to {}~productive word formation by derivational and inflectional affixations. After an overview of certain issues and relevant mathematical preliminaries, we formally present the problem and our solution. We then present results from our experiments with spelling correction in Turkish, a Ural--Altaic agglutinative language. Our results indicate that we can find the intended correct word in 95\% of the cases and offer it as the first candidate in 74\% of the cases, when the edit distance is 1.
cmp-lg/9410005
A Centering Approach to Pronouns
cmp-lg cs.CL
In this paper we present a formalization of the centering approach to modeling attentional structure in discourse and use it as the basis for an algorithm to track discourse context and bind pronouns. As described in Grosz, Joshi and Weinstein (1986), the process of centering attention on entities in the discourse gives rise to the intersentential transitional states of continuing, retaining and shifting. We propose an extension to these states which handles some additional cases of multiple ambiguous pronouns. The algorithm has been implemented in an HPSG natural language system which serves as the interface to a database query application.
cmp-lg/9410006
Evaluating Discourse Processing Algorithms
cmp-lg cs.CL
In order to take steps towards establishing a methodology for evaluating Natural Language systems, we conducted a case study. We attempt to evaluate two different approaches to anaphoric processing in discourse by comparing the accuracy and coverage of two published algorithms for finding the co-specifiers of pronouns in naturally occurring texts and dialogues. We present the quantitative results of hand-simulating these algorithms, but this analysis naturally gives rise to both a qualitative evaluation and recommendations for performing such evaluations in general. We illustrate the general difficulties encountered with quantitative evaluation. These are problems with: (a) allowing for underlying assumptions, (b) determining how to handle underspecifications, and (c) evaluating the contribution of false positives and error chaining.
cmp-lg/9410007
A Formal Look at Dependency Grammars and Phrase-Structure Grammars, with Special Consideration of Word-Order Phenomena
cmp-lg cs.CL
The central role of the lexicon in Meaning-Text Theory (MTT) and other dependency-based linguistic theories cannot be replicated in linguistic theories based on context-free grammars (CFGs). We describe Tree Adjoining Grammar (TAG) as a system that arises naturally in the process of lexicalizing CFGs. A TAG grammar can therefore be compared directly to an Meaning-Text Model (MTM). We illustrate this point by discussing the computational complexity of certain non-projective constructions, and suggest a way of incorporating locality of word-order definitions into the Surface-Syntactic Component of MTT.
cmp-lg/9410008
Recognizing Text Genres with Simple Metrics Using Discriminant Analysis
cmp-lg cs.CL
A simple method for categorizing texts into predetermined text genre categories using the statistical standard technique of discriminant analysis is demonstrated with application to the Brown corpus. Discriminant analysis makes it possible use a large number of parameters that may be specific for a certain corpus or information stream, and combine them into a small number of functions, with the parameters weighted on basis of how useful they are for discriminating text genres. An application to information retrieval is discussed.
cmp-lg/9410009
Lexical Functions and Machine Translation
cmp-lg cs.CL
This paper discusses the lexicographical concept of lexical functions and their potential exploitation in the development of a machine translation lexicon designed to handle collocations.
cmp-lg/9410010
XTAG system - A Wide Coverage Grammar for English
cmp-lg cs.CL
This paper presents the XTAG system, a grammar development tool based on the Tree Adjoining Grammar (TAG) formalism that includes a wide-coverage syntactic grammar for English. The various components of the system are discussed and preliminary evaluation results from the parsing of various corpora are given. Results from the comparison of XTAG against the IBM statistical parser and the Alvey Natural Language Tool parser are also given.
cmp-lg/9410011
Dilemma - An Instant Lexicographer
cmp-lg cs.CL
Dilemma is intended to enhance quality and increase productivity of expert human translators by presenting to the writer relevant lexical information mechanically extracted from comparable existing translations, thus replacing - or compensating for the absence of - a lexicographer and stand-by terminologist rather than the translator. Using statistics and crude surface analysis and a minimum of prior information, Dilemma identifies instances and suggests their counterparts in parallel source and target texts, on all levels down to individual words. Dilemma forms part of a tool kit for translation where focus is on text structure and over-all consistency in large text volumes rather than on framing sentences, on interaction between many actors in a large project rather than on retrieval of machine-stored data and on decision making rather than on application of given rules. In particular, the system has been tuned to the needs of the ongoing translation of European Community legislation into the languages of candidate member countries. The system has been demonstrated to and used by professional translators with promising results.
cmp-lg/9410012
Does Baum-Welch Re-estimation Help Taggers?
cmp-lg cs.CL
In part of speech tagging by Hidden Markov Model, a statistical model is used to assign grammatical categories to words in a text. Early work in the field relied on a corpus which had been tagged by a human annotator to train the model. More recently, Cutting {\it et al.} (1992) suggest that training can be achieved with a minimal lexicon and a limited amount of {\em a priori} information about probabilities, by using Baum-Welch re-estimation to automatically refine the model. In this paper, I report two experiments designed to determine how much manual training information is needed. The first experiment suggests that initial biasing of either lexical or transition probabilities is essential to achieve a good accuracy. The second experiment reveals that there are three distinct patterns of Baum-Welch re-estimation. In two of the patterns, the re-estimation ultimately reduces the accuracy of the tagging rather than improving it. The pattern which is applicable can be predicted from the quality of the initial model and the similarity between the tagged training corpus (if any) and the corpus to be tagged. Heuristics for deciding how to use re-estimation in an effective manner are given. The conclusions are broadly in agreement with those of Merialdo (1994), but give greater detail about the contributions of different parts of the model.
cmp-lg/9410013
Automatic Error Detection in Part of Speech Tagging
cmp-lg cs.CL
A technique for detecting errors made by Hidden Markov Model taggers is described, based on comparing observable values of the tagging process with a threshold. The resulting approach allows the accuracy of the tagger to be improved by accepting a lower efficiency, defined as the proportion of words which are tagged. Empirical observations are presented which demonstrate the validity of the technique and suggest how to choose an appropriate threshold.
cmp-lg/9410014
A Freely Available Syntactic Lexicon for English
cmp-lg cs.CL
This paper presents a syntactic lexicon for English that was originally derived from the Oxford Advanced Learner's Dictionary and the Oxford Dictionary of Current Idiomatic English, and then modified and augmented by hand. There are more than 37,000 syntactic entries from all 8 parts of speech. An X-windows based tool is available for maintaining the lexicon and performing searches. C and Lisp hooks are also available so that the lexicon can be easily utilized by parsers and other programs.
cmp-lg/9410015
Lexicalization and Grammar Development
cmp-lg cs.CL
In this paper we present a fully lexicalized grammar formalism as a particularly attractive framework for the specification of natural language grammars. We discuss in detail Feature-based, Lexicalized Tree Adjoining Grammars (FB-LTAGs), a representative of the class of lexicalized grammars. We illustrate the advantages of lexicalized grammars in various contexts of natural language processing, ranging from wide-coverage grammar development to parsing and machine translation. We also present a method for compact and efficient representation of lexicalized trees.
cmp-lg/9410016
Dutch Cross Serial Dependencies in HPSG
cmp-lg cs.CL
We present an analysis of Dutch cross serial dependencies in Head-driven Phrase Structure Grammar. Arguably, our analysis differs from other analyses in that we do not refer to `additional' mechanisms (e.g., sequence union, head wrapping): just standard structure sharing, an immediate dominance schema and a linear precedence rule.
cmp-lg/9410017
Concurrent Lexicalized Dependency Parsing: The ParseTalk Model
cmp-lg cs.CL
A grammar model for concurrent, object-oriented natural language parsing is introduced. Complete lexical distribution of grammatical knowledge is achieved building upon the head-oriented notions of valency and dependency, while inheritance mechanisms are used to capture lexical generalizations. The underlying concurrent computation model relies upon the actor paradigm. We consider message passing protocols for establishing dependency relations and ambiguity handling.