Although disjunction has been used in several unificationbased grammar formalisms, existing metho... more Although disjunction has been used in several unificationbased grammar formalisms, existing methods of unification have been unsatisfactory for descriptions containing large quantities of disjunction, because they require exponential time. This paper describes a method of unification by successive approximation, resulting in better average performance.
In this paper we describe a framework for research into translation that draws on a combination o... more In this paper we describe a framework for research into translation that draws on a combination of two existing and independently constructed technologies: an analysis component developed for German by the EUROTRA-D (ET-D) group of IAI and the genera-tion component developed for English by the Penman group at ISI. We present some of the linguistic impli-cations of the research and the promise it bears for furthering understanding of the translation process. 1
Nonconcatenative constraints, such as the shuffle relation, are frequently employed in grammatica... more Nonconcatenative constraints, such as the shuffle relation, are frequently employed in grammatical analyses of languages that have more flexible ordering of constituents than English. We show how it is possible to avoid searching the large space of permutations that results from a nondeterministic application of shuffle constraints. The results of our implementation demonstrate that deterministic application of shuffle constraints yields a dramatic improvement in the overall performance of a head-corner parser for German using an HPSG-style grammar. 1 Introduction Although there has been a considerable amount of research on parsing for constraint-based grammars in the HPSG (Head-driven Phrase Structure Grammar) framework, most computational implementations embody the limiting assumption that the constituents of phrases are combined only by concatenation. The few parsing algorithms that have been proposed to handle more flexible linearization constraints have not yet been applied to ...
The Penman text generation system has been used within several different experimental application... more The Penman text generation system has been used within several different experimental application domains, demonstrating that it provides the basis for an adaptable general purpose text generation capability. Linking with these applications also indicated several ways that Penman's interface with applications could be improved. Penman's interface with applications is described, focusing on SPL, a newly developed sentence plan language. SPL is a notation that can be used by text planning programs to specify plans for sentences at multiple levels of abstraction and varied amounts of detail. Sentence plans are interpreted with respect to a collection of predefined knowledge sources, thereby minimizing the size and complexity of inputs that must be dynamically constructed by the application to generate individual sentences. 1
This paper describes a particular approach to parsing that utilizes recent advances in unificatio... more This paper describes a particular approach to parsing that utilizes recent advances in unification-based parsing and in classification-based knowledge representation. As unification-based grammatical frameworks are extended to
A grammatical description often applies to a linguistic object only when that object has certain ... more A grammatical description often applies to a linguistic object only when that object has certain features. Such conditlonal descriptions can be indirectly modeled in Kay's Functional Unification Grammar (FUG) using functional descrlptions that are embedded within disjunctive alternatlves. An extension to FUG is proposed that allows for a direct representation of condltional descriptions. This extension has been used to model the input condltlons on the systems of systemlc grammar. Conditional descriptions are formally defined in terms of logical implicatlon and negatlon. This formal deftnltlon enables the use of conditional descrlptions as a general notatlonal extension to any of the unification-based gram mar representation systems currently used in computational linguistics.
Unification-based grammar formalisms use structures containing sets of features to describe lingu... more Unification-based grammar formalisms use structures containing sets of features to describe linguistic objects. Although computational algo- rithms for unification of feature structures have been worked out in experimental research, these algorithms become quite complicated, and a more precise description of feature structures is desirable. We have developed a model in which descriptions of feature structures can be regarded as logical formulas, and interpreted by sets of directed graphs which satisfy them. These graphs are, in fact, transition graphs for a special type of deterministic finite automaton.
Nonconcatenative constraints, such as the shuffle re-lation, are frequently employed in grammatic... more Nonconcatenative constraints, such as the shuffle re-lation, are frequently employed in grammatical anal-yses of languages that have more flexible ordering of constituents than English. We show how it is pos-sible to avoid searching the large space of permuta-tions that results from a nondeterministic applica-tion of shuffle constraints. The results of our imple-mentation demonstrate that deterministic applica-tion of shuffle constraints yields a dramatic improve-ment in the overall performance of a head-corner parser for German using an HPSG-style grammar. 1
We describe an approach to resolving definite descriptious and pronominal anaphora as subcases of... more We describe an approach to resolving definite descriptious and pronominal anaphora as subcases of a general strategy for presupposition satisfaction. Generally, a presupposition is satisfied in a context if the context contains a specific type of information and is organized in such a way that this information can be retrieved by the interlocutors. The model of discourse context we develop assumes that discourse structure is organized around a stack of questions under discussion, which plays a crucial role in narrowing the search for referents and other presupposed information. The algeri thmA for maintaining the discourse structures and retrieving presupposed information are presented and illustrated by several example dialogues in which human users interact with an agent tomake hotel reservatious. I I n t r o d u c t i o n "Any theory of referring expressions must take into account the discourse context in which they ot~ur. Indeed, previous research has shown that the hierarc...
When dealing with a phenomenon as vast and com plex as natural language, an experimental approach... more When dealing with a phenomenon as vast and com plex as natural language, an experimental approach is often the best way to discover new computational methods and determine their usefulness. The experimental process includes designing and selecting new experiments, carrying out the experiments, and evaluating the experiments. Most conference presentations are about finished experiments, completed theoretical results, or the evaluation of systems already in use. In this workshop setting, I would like to depart from this tendency to discuss some experiments that we are beginning to perform, and the reasons for investigating a particular approach to parsing. This approach builds on recent work in unification-based parsing and classification-based knowledge representation, developing an architecture that brings together the capabilities of these related frameworks.
Consisting of two separate papers, "Representational Issues in Systemic Functional Grammar," by C... more Consisting of two separate papers, "Representational Issues in Systemic Functional Grammar," by Christian Matthiessen and "Systemic Grammar and Functional Unification Grammar," by Robert Kasper, this document deals with systemic aspects of natural language processing and linguistic theory and with computational applications of M. A. K. Halliday's systemic functional grammar. The first paper is concerned with representation as it is defined by linguistics-how to represent language metalinguistically, or how to talk about talk. Topics covered in the paper include: (1) the lag between theory and representation; (2) dimensions of structure (including a sample grammar and discussion of the structuring, insert, expand, conflate, and preselect operators); (3) mode-of meaning and modes of structure, a characterization of representational problems of constituency, prosody, pulse, and interdependency; (4) modes of meaning and modes of choice; and (5) possible solutions. The paper also contains an appendix on representational conventions and a bibliography. The second paper looks closely into how systemic functional grammar can be represented using the framework of functional unification grammar (FUG). included are descriptions of: the ingredients of FUG (unification, paths, patterns, disjunctions, and their representations); technical details of the representation of systemic grammar in FUG; differences between FUG and systemic
International Conference on Computational Linguistics, 1988
We describe a general parsing method for systemic grammars. Systemic grammars contain a paradigma... more We describe a general parsing method for systemic grammars. Systemic grammars contain a paradigmatic analysis of language in addition to structural information, so a parser must assign a set of grammatical features and functions to each constituent in addition to producing a constituent structure. Our method constructs a parser by compiling systemic grammars into the notation of Functional Unification Grammar.
Although disjunction has been used in several unificationbased grammar formalisms, existing metho... more Although disjunction has been used in several unificationbased grammar formalisms, existing methods of unification have been unsatisfactory for descriptions containing large quantities of disjunction, because they require exponential time. This paper describes a method of unification by successive approximation, resulting in better average performance.
In this paper we describe a framework for research into translation that draws on a combination o... more In this paper we describe a framework for research into translation that draws on a combination of two existing and independently constructed technologies: an analysis component developed for German by the EUROTRA-D (ET-D) group of IAI and the genera-tion component developed for English by the Penman group at ISI. We present some of the linguistic impli-cations of the research and the promise it bears for furthering understanding of the translation process. 1
Nonconcatenative constraints, such as the shuffle relation, are frequently employed in grammatica... more Nonconcatenative constraints, such as the shuffle relation, are frequently employed in grammatical analyses of languages that have more flexible ordering of constituents than English. We show how it is possible to avoid searching the large space of permutations that results from a nondeterministic application of shuffle constraints. The results of our implementation demonstrate that deterministic application of shuffle constraints yields a dramatic improvement in the overall performance of a head-corner parser for German using an HPSG-style grammar. 1 Introduction Although there has been a considerable amount of research on parsing for constraint-based grammars in the HPSG (Head-driven Phrase Structure Grammar) framework, most computational implementations embody the limiting assumption that the constituents of phrases are combined only by concatenation. The few parsing algorithms that have been proposed to handle more flexible linearization constraints have not yet been applied to ...
The Penman text generation system has been used within several different experimental application... more The Penman text generation system has been used within several different experimental application domains, demonstrating that it provides the basis for an adaptable general purpose text generation capability. Linking with these applications also indicated several ways that Penman's interface with applications could be improved. Penman's interface with applications is described, focusing on SPL, a newly developed sentence plan language. SPL is a notation that can be used by text planning programs to specify plans for sentences at multiple levels of abstraction and varied amounts of detail. Sentence plans are interpreted with respect to a collection of predefined knowledge sources, thereby minimizing the size and complexity of inputs that must be dynamically constructed by the application to generate individual sentences. 1
This paper describes a particular approach to parsing that utilizes recent advances in unificatio... more This paper describes a particular approach to parsing that utilizes recent advances in unification-based parsing and in classification-based knowledge representation. As unification-based grammatical frameworks are extended to
A grammatical description often applies to a linguistic object only when that object has certain ... more A grammatical description often applies to a linguistic object only when that object has certain features. Such conditlonal descriptions can be indirectly modeled in Kay's Functional Unification Grammar (FUG) using functional descrlptions that are embedded within disjunctive alternatlves. An extension to FUG is proposed that allows for a direct representation of condltional descriptions. This extension has been used to model the input condltlons on the systems of systemlc grammar. Conditional descriptions are formally defined in terms of logical implicatlon and negatlon. This formal deftnltlon enables the use of conditional descrlptions as a general notatlonal extension to any of the unification-based gram mar representation systems currently used in computational linguistics.
Unification-based grammar formalisms use structures containing sets of features to describe lingu... more Unification-based grammar formalisms use structures containing sets of features to describe linguistic objects. Although computational algo- rithms for unification of feature structures have been worked out in experimental research, these algorithms become quite complicated, and a more precise description of feature structures is desirable. We have developed a model in which descriptions of feature structures can be regarded as logical formulas, and interpreted by sets of directed graphs which satisfy them. These graphs are, in fact, transition graphs for a special type of deterministic finite automaton.
Nonconcatenative constraints, such as the shuffle re-lation, are frequently employed in grammatic... more Nonconcatenative constraints, such as the shuffle re-lation, are frequently employed in grammatical anal-yses of languages that have more flexible ordering of constituents than English. We show how it is pos-sible to avoid searching the large space of permuta-tions that results from a nondeterministic applica-tion of shuffle constraints. The results of our imple-mentation demonstrate that deterministic applica-tion of shuffle constraints yields a dramatic improve-ment in the overall performance of a head-corner parser for German using an HPSG-style grammar. 1
We describe an approach to resolving definite descriptious and pronominal anaphora as subcases of... more We describe an approach to resolving definite descriptious and pronominal anaphora as subcases of a general strategy for presupposition satisfaction. Generally, a presupposition is satisfied in a context if the context contains a specific type of information and is organized in such a way that this information can be retrieved by the interlocutors. The model of discourse context we develop assumes that discourse structure is organized around a stack of questions under discussion, which plays a crucial role in narrowing the search for referents and other presupposed information. The algeri thmA for maintaining the discourse structures and retrieving presupposed information are presented and illustrated by several example dialogues in which human users interact with an agent tomake hotel reservatious. I I n t r o d u c t i o n "Any theory of referring expressions must take into account the discourse context in which they ot~ur. Indeed, previous research has shown that the hierarc...
When dealing with a phenomenon as vast and com plex as natural language, an experimental approach... more When dealing with a phenomenon as vast and com plex as natural language, an experimental approach is often the best way to discover new computational methods and determine their usefulness. The experimental process includes designing and selecting new experiments, carrying out the experiments, and evaluating the experiments. Most conference presentations are about finished experiments, completed theoretical results, or the evaluation of systems already in use. In this workshop setting, I would like to depart from this tendency to discuss some experiments that we are beginning to perform, and the reasons for investigating a particular approach to parsing. This approach builds on recent work in unification-based parsing and classification-based knowledge representation, developing an architecture that brings together the capabilities of these related frameworks.
Consisting of two separate papers, "Representational Issues in Systemic Functional Grammar," by C... more Consisting of two separate papers, "Representational Issues in Systemic Functional Grammar," by Christian Matthiessen and "Systemic Grammar and Functional Unification Grammar," by Robert Kasper, this document deals with systemic aspects of natural language processing and linguistic theory and with computational applications of M. A. K. Halliday's systemic functional grammar. The first paper is concerned with representation as it is defined by linguistics-how to represent language metalinguistically, or how to talk about talk. Topics covered in the paper include: (1) the lag between theory and representation; (2) dimensions of structure (including a sample grammar and discussion of the structuring, insert, expand, conflate, and preselect operators); (3) mode-of meaning and modes of structure, a characterization of representational problems of constituency, prosody, pulse, and interdependency; (4) modes of meaning and modes of choice; and (5) possible solutions. The paper also contains an appendix on representational conventions and a bibliography. The second paper looks closely into how systemic functional grammar can be represented using the framework of functional unification grammar (FUG). included are descriptions of: the ingredients of FUG (unification, paths, patterns, disjunctions, and their representations); technical details of the representation of systemic grammar in FUG; differences between FUG and systemic
International Conference on Computational Linguistics, 1988
We describe a general parsing method for systemic grammars. Systemic grammars contain a paradigma... more We describe a general parsing method for systemic grammars. Systemic grammars contain a paradigmatic analysis of language in addition to structural information, so a parser must assign a set of grammatical features and functions to each constituent in addition to producing a constituent structure. Our method constructs a parser by compiling systemic grammars into the notation of Functional Unification Grammar.
Uploads
Papers by Robert Kasper