Abstract. This paper describes an approach for handling structural divergences and recovering dropped arguments in an implemented Korean to English machine translation system. The approach relies on canonical predicate-argument structures (or dependency structures), which p r o vide a suitable pivot representation for the handling of structural divergences and the recovery of dropped arguments. It can also be converted to and from the interface representations of many o -the-shelf parsers and generators.
This paper describes the initial results of an experiment in integrating knowledge-based text processing with real-world reasoning in a question answering system. Our MOQA "meaning-oriented question answering" system seeks answers to questions not in open text but rather in a structured fact repository whose elements are instances of ontological concepts extracted from the text meaning representations (TMRs) produced by the OntoSem text analyzer. The query interpretation and answer content formulation modules of MOQA use the same knowledge representation substrate and the same static knowledge resources as the ontological semantic (OntoSem) semantic text analyzer. The same analyzer is used for deriving the meaning of questions and of texts from which the fact repository content is extracted. Inference processes in question answering rely on ontological scripts (complex events) that also support reasoning for purely NLP-related purposes, such as ambiguity resolution in its many guises.
With the emergence of object-oriented technology and user-centered software engineering paradigms, the requirements analysis phase has changed in two important ways: it has become an iterative activity, and it has become more closely linked to the design phase of software engineering (Davis, 1993). A requirements analyst builds a formal object-oriented (OO) domain model. A user (domain expert) validates the domain model. The domain model undergoes subsequent evolution (modification or adjustment) by a (perhaps different) analyst. Finally, the domain model is passed to the designer (system analyst), who refines the model into a OO design model used as the basis for implementation. Thus, we can see that the OO models form the basis of many important flows of information in OO software engineering methodologies. How can this information best be communicated? It is widely believed that graphical representations are easy to learn and use, both for modeling and for communication among the engineers and domain experts who tqgether develop the OO domain model. This belief is reflected by the large number of graphical OO modeling tools currently in research labs and on the market. However, this belief is not accurate, as some recent empirical studies show. For example, Kim (1990) simulated a modeling task with experienced analysts and a validation task with sophisticated users not familiar with the particular graphical language. Both user groups showed semantic error rates between 25% and 70% for the separately scored areas of entities, attributes, and relations. Relations were particularly troublesome to both analysts and users. Petre (1995) compares diagrams with textual representations of nested conditional structures (which can be compared to OO modeling in the complexity of the "paths" through the system). She finds that "the intrinsic difficulty of
In this paper we describe an implemented framework for developing monolingual or multilingual natural language generation (NLG) applications and machine translation (MT) applications. The framework demonstrates a uniform approach to generation and transfer based on declarative lexico-structural transformations of dependency structures of syntactic or conceptual levels ("uniform lexico-structural processing").We describe how this framework has been used in practical NLG and MT applications, and report the lessons learned.
We describe the design of an MT system that employs transfer rules induced from parsed bitexts and present evaluation results. The system learns lexico-structural transfer rules using syntactic pattern matching, statistical co-occurrence and errordriven filtering. In an experiment with domainspecific Korean to English translation, the approach yielded substantial improvements over three baseline systems.
This paper describes a novel approach to inducing lexico-structural transfer rules from parsed bi-texts using syntactic pattern matching, statistical cooccurrence and error-driven filtering. We present initial evaluation results and discuss future directions.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.