This paper describes techniques for automatic construction of dictionaries for use in large-scale foreign language tutoring (FLT) and interlingual machine translation (MT) systems. The dictionaries are based on a language-independent representation called lexical conceptual structure (LCS). A primary goal of the LCS research is to demonstrate that synonymous verb senses share distributional patterns. In this paper, we show how the syntax-semantics relation can be used to develop a lexical acquisition approach that contributes both toward the enrichment of existing online resources and toward the development of lexicons containing more complete information than is provided in any of these resources alone. We start by describing the structure of the LCS and showing how this representation is used in FLT and MT. We then focus on the problem of building LCS dictionaries for large-scale FLT and MT. First, we describe authoring tools for manual and semi-automatic construction of LCS dictionaries; we then present a more sophisticated approach that uses linguistic techniques for building word defmitions automatically. These techniques have been implemented as part of a set of lexicon-development tools used in the MILT FLT project (Dorr et al., 1995; Sams, 1995; Weinberg et al., 1995) and in the PRINCITRAN MT project (Dorr et al., 1995b).

Verbal and compositional lexical aspect provide the underlying temporal structure of events. Knowledge of lexical aspect, e.g.. (a)telicity, is therefore required for interpreting event sequences in dis- course (Dowry, 1986; Moens and Steed- man, 1988; Passoneau, 1988), interfacing to temporal databases (Androutsopoulos, 1996), processing temporal modifiers (Antonisse, 1994), describing allowable alternations and their semantic effects (Resnik, 1996; Tenny, 1994), and selecting tense and lexical items for natural language generation ((Dorr and Olsen, 1996; Klavans and Chodorow, 1992), cf. (Slobin and Bocaz, 1988)). We show that it is possible to represent lexical aspect--both verbal and compositional--on a large scale, using Lexical Conceptual Structure (LCS) representations of verbs in the classes cat- aloged by Levin (1993). Ve show how proper consideration of these universal pieces of verb meaning may be used to refine lexical representations and derive a range of meanings from combinations of LCS representations. A single algorithm may therefore be used to determine lexical aspect classes and features at both verbal and sentence levels. Finally, we illustrate how knowledge of lexical aspect facilitates the interpretation of events in NLP appli- cations.

This paper is a survey of the current machine translation research in the US, Europe and Japan. A short history of machine translation is presented first, followed by an overview of the current research work. Representative examples of a wide range of different approaches adopted by machine translation researchers are presented. These are described in detail along with a discussion of the practicalities of scaling up these approaches for operational environments. In support of this discussion, issues in, and techniques for, evaluating machine translation systems are addressed.

tween the words in a sentence. PRINCIPAR attempts to construct a parse for the full sentence. However, when it fails to do that, it retrieves parse fragments that cover the complete sentence. Information is extracted from the dependency trees by a subtree pattern matcher. The format of NE and CO outputs do not meet the standard of the scoring software because of arbitrary insertion and deletion of white spaces. A separate program is used to resolved the differences. The PIE system is implemented in about 38k lines of C++, about 33k to 34k lines were written before MUC-6. It contains an interpreter for LISP-like expressions so that all the knowledge structures, such as finite automata for finding sentence boundaries, the lexicon, the lexical rules, the grammar network, and extraction rules, are written in LISP-like expressions. Performance in MUC-6 Considering the very limited time and human resources with which the PIE system was developed, it did very well in all of

This paper describes a multi-level design, i.e., a non-uniform approach to interlingual machine translation (MT), in which distinct representational languages are used for different types of knowledge. We demonstrate that a linguistically-motivated "division of labor" across multiple representation levels has not complicated, but rather has readily facilitated, the identification and construction of systematic relations at the interface between each level. Our approach assumes an interlingua derived from the lexical semantics and predicate decomposition approaches of Jackendoff (1983; 1990) and Levin and Rappaport-Hovav (1995a; 1995b). We describe a model of interpretation and representation of natural language sentences which has been implemented as part of an interlingual MT system called PRINCITRAN.

: Verbal and compositional lexical aspect provide the underlying temporal structure of events. Knowledge of lexical aspect, e.g., (a)telicity, is therefore required for interpreting event sequences in discourse (Dowty, 1986; Moens and Steedman, 1988; Passoneau, 1988), interfacing to temporal databases (Androutsopoulos, 1996), processing temporal modifiers (Antonisse, 1994), describing allowable alternations and their semantic effects (Resnik, 1996; Tenny, 1994), and selecting tense and lexical items for natural language generation ((Dorr and Olsen, 1996; Klavans and Chodorow, 1992), cf. (Slobin and Bocaz, 1988)). We show that it is possible to represent lexical aspect---both verbal and compositional---on a large scale, using Lexical Conceptual Structure (LCS) representations of verbs in the classes cataloged by Levin (1993). We show how proper consideration of these universal pieces of verb meaning may be used to refine lexical representations and derive a range of meanings from combin...

This paper describes how the most complete recent linguistic results on Korean scrambling (switching of word order) can be readily incorporated into an existing principlesand -parameters parser with minimal additional machinery. Out of all 29 sets of examples in chapters 2.2 and 3.2 of perhaps the most advanced linguistic analysis of Korean scrambling, [5], 26 sets of examples can be correctly parsed, greatly extending the variety of scrambling handled by any current parser. This approach is compared to other current approaches to scrambling, such as PRINCIPAR and Tree Adjoining Grammar. Thesis Supervisor: Robert C. Berwick Title: Professor of Computer Science and Engineering and Computational Linguistics Acknowledgments Firstly, the author would like to thank his family and friends for all their support over the years. The author also would like to thank Prof. Robert C. Berwick, Dr. Sandiway Fong, Dr. Youngsuk Lee and Dr. Dekang Lin for their generous help. To My Parents Contents...