sues in MT and the stands on these issues taken in the KBMT approach. The most important is the issue of transfer versus interlingua. In a nutshell, the question is: can analyzers and generators for the various languages in an MT system share the same (interlingual) representation, or must representations vary, with transfer modules being responsible for converting from the source language representation to the target language representation ? I think the authors are correct in concluding that this controversy really boils down to the question of how much semantic analysis is performed in an MT system. The more semantic analysis, the more like an interlingua the representation is likely to be. Aside from this point, though, ! did not find the discussion to be particu- 207 Computational Linguistics Volume 19, Number 1 larly original; many of the same points have been raised countless times before, dating as far back as Bar-Hillel's 1960 article discussing the feasibility of fully automa

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.

This dissertation is an investigation into how ambiguity should be classified for authors and readers of text, and how this process can be automated. Usually, authors and readers disambiguate ambiguity, either consciously or unconsciously. However, disambiguation is not always appropriate. For instance, a linguistic construction may be read differently by different people, with no consensus about which reading is the intended one. This is particularly dangerous if they do not realise that other readings are possible. Misunderstandings may then occur. This is particularly serious in the field of requirements engineering. If requirements are misunderstood, systems may be built incorrectly, and this can prove very costly. Our research uses natural language processing techniques to address ambiguity in requirements. We develop a model of ambiguity, and a method of applying it, which represent a novel approach to the problem described here. Our model is based on the notion that human perception is the only valid criterion for judging ambiguity. If people perceive very differently how an ambiguity should be read, it will cause misunderstandings. Assigning a preferred reading to it is therefore unwise. In

Abstract not found

This discussion outlines and implements the theory of an inductive inference technique that automatically discovers classes among large numbers of input patterns, generates operational definitions of class membership with explicit levels of confidence, creates a continuously updated "self-organized" coded hierarchical taxonomic classification of patterns, and recognizes to which already discovered class or classes, if any, a new input belongs in an information-theoretically efficient way. Relationships to the "scientific method" and learning are discussed.

A unification-based grammar is a type of language description well suited for the implementation on a computer. As many other contemporary grammar theories, almost all the unification-based ones advocate the usage of a set of phrase structure rules to describe how acceptable utterances are formed in a particular language. The constituents of the rules are annotated with a set of features. Rules may be applied to phrases only if the structure of a rule and a phrase match each other, or in other words, if they unify. That is, if their features only take on values of the same types and within the same ranges. The thesis describes the implementation of a large-scale unification-based grammar for Swedish. Part of the text is devoted to the more theoretical and linguistic sides, discussing all the different grammar rules, while concentrating the main effort on the treatment of verb phrases. Another part addresses a range of different application areas in which a language processing system eq...

Abstract not found

The recent report for the Commission of the European Communities on current multilingual activities in the field of scientific and technical information 1 and the 1977 conference on the same theme 2 both included substantial sections on operational and experimental machine translation systems, and in its Plan of action 3 the Commission announced its intention to introduce an operational machine translation system into its departments and to support research projects on machine translation. This revival of interest in machine translation may well have surprised many who have tended in recent years to dismiss it as one of the 'great failures ' of scientific research. What has changed? What grounds are there now for optimism about machine translation? Or is it still a 'utopian dream'? The aim of this review is to give a general picture of present activities which may help readers to reach their own conclusions. After a sketch of the historical background and general aims (section I), it describes operational and experimental machine translation systems of recent years (section II), it continues with descriptions of interactive (man-machine) systems and machine-assisted translation (section III), and it concludes with a general survey of present problems and future possibilities (section IV).

While recent statistical methods for disambiguation in natural language processing have been very successful, earlier arguments for the position that fully successful disambiguation requires reasoning to new conclusions from old facts still hold. We explore ways of complementing statistical approaches with the use of ‘domain theories’, collections of facts and axioms that characterise the typical structure of some task domain. In particular, we hypothesise that disambiguation decisions can supply tacit information about such theories, and that the theories can be in part automatically induced from such data. We describe a pilot experiment in which a partial domain theory for the domain of air travel information was induced from a corpus of disambiguated example sentences, the resulting theory then being used successfully in disambiguating other sentences from the same domain.

The history of machine translation is described from its beginnings in the 1940s to the present day. In the earliest years, efforts were concentrated either on developing immediately useful systems, however crude in their translation quality, or on fundamental research for high quality translation systems. After the ALPAC report in 1966, which virtually ended MT research in the US for more than a decade, research focussed on the development of systems requiring human assistance for producing translations of technical documentation, on translation tools for direct use by translators themselves, and, in recent years, on systems for translating email, Web pages and other Internet documentation, where poor quality is acceptable in the interest of rapid results.