Abstract. In the last ten years there has been a significant amount of research in Machine Translation within a “new ” paradigm of empirical approaches, often labelled collectively as “Example-based” approaches. The first manifestation of this approach caused some surprise and hostility among observers more used to different ways of working, but the techniques were quickly adopted and adapted by many researchers, often creating hybrid systems. This paper reviews the various research efforts within this paradigm reported to date, and attempts a categorisation of different manifestations of the general approach.

The Pangloss Example-tased Machine Translation engine (PanEBMT) is a translation system requiring essentially no knowledge of the structure of a language, merely a large parallel corpus of example sentences and a bilingual dictionary. Input

Example-Based Machine Translation (EBMT) using partial exact matching against a database of translation examples has proven quite successful, but requires a large amount of pre-translated text in order to achieve broad coverage of unrestricted text. By adding linguistically tagged entries to the example base and permitting recursive matches that replace the matched text with the associated tag, substantial reductions in the required amount of pre-translated text can be achieved. A modest investment of time -- on the order of two person-weeks -- adding linguistic knowledge reduces the required example text by a factor of six or more, while retaining comparable translation quality. This reduction makes EBMT more attractive for so-called "low-density" languages for which little data is available.

An Example-Based Machine Translation system is supplied with a sentence-aligned bilingual corpus, but no other knowledge sources. Using the knowledge implicit in the corpus, it generates a bilingual word-for-word dictionary for alignment during translation. With such an automatically-generated dictionary, the system covers (with equivalent quality) more of its input on unseen texts than the same system does when provided with a manually-created general-purpose dictionary and other knowledge sources.

The transfer phase in machine translation (MT) systems has been considered to be more complicated than analysis and generation, since it is inherently a conglomeration of individual lexical rules. Currently sane attempts are being made to use casebased reasoning in machine translation, that is, to make decisions on the basis of translation examples at apprc priate points in MT. This paper proposes a new type of transfer system, called a Similarity-driven 71'aris- re,' Sgstem (Sim7'an), fo .... in sue[ .... based MT (CBMT).

This paper describes an experiment in examplebased machine translation (EBMT) on full text. The unit of translation is a text chunk of arbitrary length, in contrast to sentence-level EBMT experiments. Intra- and inter-language matching techniques and metrics used in the experiment are described. Keywords: Example-based MT, Corpus-based NLP Introduction The growth rate of theoretical studies of language structure and use stubbornly remains higher than the improvement rate of large-scale applications. It has been repeatedly proved that large-scale realistic NLP applications carry a prohibitive price tag of large-scale, routine acquisition of knowledge about language and about the world, collected in computational grammars, lexicons and domain models. Strategically, there are several ways of dealing with this problem: ffl biting the bullet and going through a massive knowledge acquisition effort, either generalpurpose (e.g., the CYC project, Lenat et al., 1990) or domain-specific (e.g....

This paper describes two approaches to matching input strings with strings from a translation archive in the example-based machine translation paradigm- the more canonical "chunking + matching + recombination " method and an alternative method of matching at the level of complete sentences. The latter produces less exact matches while the former suffers from (often serious) translation quality lapses at the boundaries of recombined chunks. A set of text matching criteria was selected to reflect the trade-off between utility and computational price of each criterion. A metric for comparing text passages was devised and calibrated with the help of a specially constructed diagnostic example set. A partitioning algorithm was developed for finding an optimum "cover " of an input string by a set of best-matching shorter chunks. The results were evaluated in a monolingual setting using an existing MT post-editing tool: the distance between the input and its best match in the archive was calculated in terms of the number of keystrokes necessary to reduce the latter to the former. As a result, the metric was adjusted and an experiment was run to test the two EBMT methods, both on the training corpus and on the working corpus (or "archive") of some 6,500 sentences. 47 The growth rate of theoretical studies of language structure and use stubbornly remains higher than the improvement

We present a set of algorithms that enable us to translate natural language sentences by exploiting both a translation memory and a statistical-based translation model. Our results show that an automatically derived translation memory can be used within a statistical framework to often find translations of higher probability than those found using solely a statistical model.

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.

The example-based approach to MT is becoming increasingly popular. However, such is the variety of techniques and methods used that it is difficult to discern the overall conception of what examplebased machine translation (EBMT) is and/or what its practitioners conceive it to be. Although definitions of MT systems are notoriously complex, an attempt is made to define EBMT in contrast to other MT architectures (RBMT and SMT).