Abstract not found

Abstract not found

This paper shows how DATR, a widely used formal language for lexical knowledge representation, can be used to define an [TAG lexicon as an inheritance hierarchy with internal lexical rules. A bottom-up learural encoding is used for [TAG trees and this allows lexical rules to be implemen- ted as covariation constraints within fea- ture structures. Such an approach eliminates the considerable redundancy otherwise associated with an I_TAG lexicon.

In this paper we view Lexicalized Tree Adjoining Grammars as the compilation of a more abstract and modular layer of linguistic description :the metagrammar (MG). MG provides a hierarchical representation of lexicosyntactic descriptions and principles that capture the well-formedness of lexicalized structures, expressed using syntactic functions. This makes it possible for a tool to compile an instance of MG into an LTAG, automatically performing the relevant combinations of linguistic phenomena. We then describe the instantiation of an MG for Italian and French. The work for French was performed starting with an existing LTAG, which has been augmented as a result. The work for Italian was performed by systematic contrast with the French MG. The automatic compilation gives two parallel LTAG, compatible for multilingual NLP applications.

In this paper, we propose to introduce syntactic classes in a lexicalized formalism like dependency. Subcategories of words are organized hierarchically from a general, abstract level (syntactic categories) to a word-specific level (single lexical items). The formalism is linguistically perspicuous, and useful for processing. A parsing model is defined, that uses the hierarchical mixed-grain representation (compiled in a tabular form), thus benefitting from both general and specific information in different stages of processing. 1 Introduction Much recent work in linguistics and computational linguistics emphasizes the role of lexical information in syntactic representation and processing. This emphasis given to the lexicon is the result of a gradual process. The original trend in computational linguistics has been to individuate categories of words having related characteristics -- the traditional syntactic categories like verb, noun, adjective, etc. -- and to express, by means of ...

The question we address in this paper is whether `Lexical Rules' deserve their grand status, a status that is often conveyed by a special purpose formalism and/or a separate component, one that may even be external to the lexicon proper. We will argue that they do not and that a lexical knowledge representation language that is as expressive as it needs to be for other lexical purposes will, ipso facto, be expressive enough to encode `Lexical Rules' internally as lexical rules. Such internal encoding is not only possible but also desirable since `Lexical Rules' will then automatically acquire other characteristics which are now standardly associated with common or garden lexical rules, including inheritance, generalization by default, and the ability to relate lexical information from different levels of linguistic description. We give examples of what we take to be instances of common or garden lexical rules and then show how the same formal machinery provides for the statement of a P...

This paper describes the use of verb class memberships as a means of capturing generalizations about manner-of-motion verbs in Synchronous Tree Adjoining Grammars, STAGs, [20, 21, 22]. This approach allows STAGs, which are essentially transfer-based, to take advantage of the same types of generalizations which are generally thought of as wholly the domain of interlingua systems- without giving up any of the lexical specificity unique to transfer-based systems. In this way a machine translation system based on STAGs can respond with seamless flexibility to a wide spectrum of phenomena being presented for translation ranging from idioms and idiosyncratic lexical items to well-behaved verbs that follow lexical rules. 1

We discuss a grammar development process used to generate the trees of the wide-coverage Lexicalized Tree Adjoining Grammar (LTAG) for English of the XTAG Project. Result of the coupling of Becker's metarules and a simple yet principled hierarchy of rule application, the approach has been successful to generate the large set of verb trees in the grammar, from a very small initial set of trees.

This paper addresses various issues related to representing the Korean language using Tree Adjoining Grammars. Topics covered include Korean grammar using TAGs, Machine Translation between Korean and English using Synchronous Tree Adjoining Grammars (STAGs), handling scrambling using Multi Component TAGs(MC-TAGs), and recovering empty arguments. The data for the parsing is from US military telecommunication messages. Chapter 1 Overview Linguistics and computer science are merging closer together in natural language processing. Especially in Korean, many scholars have made notable contributions including a wide spectrum of studies such as pragmatics of compound verbs, indirect speech acts, and parsing methods. Unfortunately, most of them paid little attention to combining these ideas together to formalize linguistic theory into computer science, and vice versa. Beyond just observing a phenomenon, linguists should be able to formalize it, or give a reason to prove that their observatio...

This paper describes the use of a compact encoding scheme to represent the trees of the wide-coverage DTG grammar currently being developed in the LEXSYS project (Caroll et al 1998). The encoding scheme is derived from the scheme for LTAG grammars described in Evans, Gazdar and Weir (1995), but the LEXSYS grammar is the first attempt to apply these ideas on a larger scale. In this paper we report on the approach taken and discuss some technical improvements to the encoding scheme that we have introduced to overcome problems of scaling.