M. Nagao. 1984. A framework of a mechanical translation between Japanese and English by analogy principle. In: A.Elithorn and R.Banerji (eds.) Artificial and human intelligence (Amsterdam: North-Holland), 173-180.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....away from empiricism in linguistics under the in uence of Chomsky. Successes in speech recognition [9] as well as increased computer power and growing amounts of text available in electronic form motivated their reintroduction. The earliest instance was example based MT, proposed by Nagao in 1984 [97]. Although example based MT received considerable attention and funding, especially in Japan, the greatest excitement and controversy was generated by the statistical approach pioneered by a group of former speechrecognition researchers at IBM [23] in the late 1980 s and early 90 s (see appendix ....

Makoto Nagao. A framework of a mechanical translation between Japanese and English by analogy principle. In A. Elithorn and R. Banerji, editors, Arti cial and human intelligence. Elsevier, 1984.

....to produce the translation result shown in Figure 4. Analysis application of analysis kn0wlege Input [ Transfer application of information transfer [ knowlege output Figure 4 Relation between transfer and analysis At present, we are providing analysis knowledge for normalization [Nagao 84] and for structuring with TDMT. In the following sections we will explain the cooperation mechanism between transfer and analysis based on these two kinds of analysis knowledge. 4.2.1 Analysis knowledge for normalization Normalization is putting together minor colloquial expressions into ....

....flexibly communicates various kinds of knowledge including context md generation knowledge, and various kinds of frameworks such as a rule based and a statistical framework are useful to improve the translation performance. 7 Related Research The example based approach was advocated by Nagao [Nagao 84] The essence of this approach is (a) retrieval of similar examples from a bilingual database and (b) applying the examples to translate the input. Other research has emerged following this line, including EBMT [Sumira and Iida 91] MBT [Sato and Nagao 901, and ABMT [Sadler 89] EBMT uses phrase ....

Nagao, M.: A framework of a mechanical translation between Japanese and English by analogy principle, in Artificial and Human Intelligence, ed. Elithom, A. and Banerji, R.. North-Holland, pp.173180, (1984).

....but it is quite difficult because the trouble of description and the cost of calculation. 2. 2 Example Based Translation Besides the conventionM translation method above, a machine translation system based on translation examples (pairs of source texts and their translations) is also proposed [Nagao 84, ato 90, Soreits 90] This type of system, called Example Based Machine Translation, has stored a large amount of bilingual translation exampigs as a database, and translates input exprgssions by retrieving an example most similar to the input from the database. There is no failure of output in ....

M.Nagao "A Framework of a Mechani- cal Translation between Japanese and English by Analogy Principle" Ar*iticiM and Human Intelligeace, ed. A.Elithorn arid R. Banerji, NorthsHolland , 1984

....CTRA: Compound word translation DTRA: Data production for STRA EXTRA: Fixed sentence extraction An overview of ENTS Currently he is working at ATR Interpreting Telecom munications Research Laboratories. 28 Recently, several example based MTs were proposed for processing fixed expressions [Nagao84][Sumita91 ] Furuse proposed a cooperative method using tightly woven combination of example and rule based approaches I Furuse92] In contrast to their approach, we use the two methods independently. Therefore, the translation accuracy of our example based method is guaranteed to be 100 . ....

Nagao, M.: "A Framework of a Mechanical Translation between Japanese and English by Analogy Principle", in Elithorn, A. and R. Bernerji (eds.) Artificial and Human Intelligence, North-Holland, pp. 173-180 (1984).

....module which applies the transfer knowledge to each input sentence. Other modules, such as lexical processing, analysis, and generation, cooperate with the transfer module to improve translation performance. With this transfer centered translation mechanism together with the examplebased fYamework[3], 4] which conducts distance calculations between linguistic expressions using the semantic hierarchy, TDMT performs efficient and robust translation. TDMT is especially useful for spoken language translation, since spoken language expressions tend to deviate from conventional grammars and ....

Nagao, M.: "A Framework of a Mechanical Translation Between Japanese and English by Analogy Principle," in Artificial and Human Intelligence, Elithorn, A. and Banerji, R. (eds.), North-Holland, pp.173-180 (1984) 114] Sato, S. and Nagao M.: "Toward Memory- Based Translation," Proc. of COLING-90

....of clustering to make retrieval of the best matching example from the database more efficient. Results on a large number of test cases from the CELEX database are presented. 1. 1NTRODUCTION EBMT is based on the idea of performing translation by imitating translation examples of similar sentences [Nagao 84] In this type of translation system, a large amount of bi multi lingual translation examples has been stored in a textnal database and input expressions are rendered in the target language by retrieving from the database that example which is most similar to the input. There are three key ....

Nagao M., (1984). "A framework of a mechanical translation between Japanese and English by analogy principle". Artificial and Human Intelligence, ed. Elithorn A. and Banerji R., Northtlolland, pp 173-180.

.... the most probable words in the target language, EBMT uses pattern matching techniques to translate subparts of the given input [1] Also, Siskind has dealt with the question of what matches what in the examples in some detail for the case of lexical matching [12] EBMT has been proposed by Nagao [7] as Translation by Analogy which is in parallel with memory based reasoning [13] case based reasoning [9] and derivational analogy [2] The example based translation relies on the use of past translation examples to derive a translation for a given input. The input and the example translations ....

Nagao, M. A.: Framework of a Mechanical Translation between Japanese and English by Analogy Principle (1985).

....the sentences (parse) before acquiring transfer rules (cf. 12] 9] 10] 13] and [11] This approach has the advantage of acquiring structural as well as lexical correspondences. A large, syntactically analyzed, aligned corpus may serve as an example base for a form of example based MT (cf. [15], 16] 10] and [7] This paper describes: 1) an efficient algorithm for deriving an alignment for a pair of source and target language parse trees; and (2) a procedure for deriving transfer rules from this alignment. Each transfer rule consists of a pair of tree fragments derived by cutting ....

Makao Nagao. A Framework of a Mechanical Translation between Japanese and English by Analogy Principle. In Alick Elithorn and Ranan Banerji, editors, Artificial and Human Intelligence. Elsevier Science Publishers B.V., Amsterdam, 1984.

....to statistical training and extraction of more symbolic approaches. This intersection of statistical and symbolic paradigms is relevant to the hybrid techniques discussed in Section 5.3 below. 5.2. 2 Example (or Case Memory ) Based MT (EBMT) Example Based MT (EBMT) first suggested by Nagao [146], emulates human translation practice in recognizing the similarity of a new source language sentence or phrase to a previously translated item and using this previous translation to perform Translation by Analogy . Sato and Nagao [181] implemented an experimental EBMT system to demonstrate the ....

.... 40 Jones and Tsujii [111] DBMT Kaplan et al. 115] CBMT, RBMT Kaplan and Wedekind [116] CBMT, RBMT Komatsu et al. 127] SBMT Landsbergen et al. 129] RBMT Maruyama and Watanabe [138] EBMT Maruyama et al. [137] SBMT McCord [140] RBMT McLean [141] EBMT, NBMT Mitamura et al. 159] KBMT Nagao [146] EBMT Nagao [148] KBMT Nirenburg et al. 155] KBMT Nirenburg et al. 154] EBMT Nomiyama [158] EBMT, SBMT Okumura et al. 162] EBMT, RBMT Richardson et al. 173] EBMT Sadler and Arnold [176] CBMT Saito and Tomita [178] DBMT Sanfilippo et al. 179] LBMT Sato [180] EBMT Sato and Nagao ....

M. Nagao. A Framework of a Mechanical Translation Between Japanese and English By Analogy Principle. In A. Elithorn and R. Banerji, editor, Artificial and Human Intelligence. North-Holland, 1984.

.... pairs (Japanese English) 1 Introduction Obtaining aligned pairs of source and target language sentences is an important precursor to further work in many natural language processing applications, including corpus trained (e.g. 9] 12] 17] 1] 8] 3] 20] 21] and examplebased (e.g. [22], 23] 6] approaches to machine translation. Human translators may also use aligned bitexts to aid them in translating new similar texts. Even if an identical sentence has not been previously translated, translations of similar sentences may be used as models. Manual alignment unduly ....

Makao Nagao. A Framework of a Mechanical Translation between Japanese and English by Analogy Principle. In Alick Elithorn and Ranan Banerji, editors, Artificial and Human Intelligence. Elsevier Science Publishers B.V., Amsterdam, 1984.

....to statistical training and extraction of more symbolic approaches. This intersection of statistical and symbolic paradigms is relevant to the hybrid techniques discussed in Section 5.3 below. 5.2. 2 Example (or Case Memory ) Based MT (EBMT) Example Based MT (EBMT) first suggested by Nagao [146], emulates human translation practice in recognizing the similarity of a new source lan 33 guage sentence or phrase to a previously translated item and using this previous translation to perform Translation by Analogy . Sato and Nagao [183] implemented an experimental EBMT system to demonstrate ....

M. Nagao. A Framework of a Mechanical Translation Between Japanese and English By Analogy Principle. In A. Elithorn and R. Banerji, editor, Artificial and Human Intelligence. North-Holland, 1984.

.... [1 31215 8] 53[5 8] Products [1 31410 91 6110 91 Part of a living thing [1 3157 621 Plant [1 3155 631 Nature [1 3152 641 Location [1 3] 17 657 Quaatity [1 3119 711 Time [1 3] 16 811 Phenomenon [1 3151011 911121 Abstract relation [1 31110 58] aa[0 58] Human activity [1 3158, 1 31310 8] ab[0 9] Other 4 d the is a hierarchy. The top five levels are expressed by the first five digits, the sixth level is expressed by the next two digits, and the last level is expressed by the last three digits. Although we have used BGH, Msort can also be used with other thesauri in other languages. ....

....4, 11] and the investigation of corpus based learning algorithms is attracting much attention [8] In this section, we demonstrate how an Msort can be useful in the construction of corpora. Suppose that we want to disambiguate the meanings of of in NP X of NP Y by using the example based method [9]. In this case, we need a tagged corpus Table 9. Example construction of a case frame of the verb eat (a) Results of an Msort of terms in the nominative case (Animal) cattle a calf fish a (Human) we, us, all, myself, babies, a parent, a sister, a customer, Japanese, a nurse, a writer (b) ....

Makoto Nagao. A Framework of a Mechanical Translation between Japanese and English by Analogy Principle. Artificial and Human Intelligence, pages 173 180, 1984.

....of the conducted experiments. Finally, Section 5 concludes the paper. 2 Related work The example(s) an EBMT system determines to be equivalent (or at least similar enough) to the text to be translated varies according to the approach taken by the system. The pioneer in EBMT system is Nagao [6] who suggested to emulate human translation practice in recognizing the similarity of a new source language sentence to a previously translated item by selecting identical phrases available in the translation memory except for a similar content word. This approach has been further developed by ....

M. Nagao. A Framework of a Mechanical Translation between Japanese and English by Analogy Principle. Nato Publications, 1984.

....a wildcard, in which case it can be matched to any value. The system systematically generates all combinations of wildcards, from fewest to greatest, until a set of 100 similar clauses are found. The extraction of similar sentences from a corpus is also used in example based machine translation [17]; the translation of a sentence is guided by the translation of a similar sentence in a bilingual corpus. 3.2 Aspectual Indicators Once 100 similar clauses have been extracted, the occurrences of aspectual markers can be examined to predict the aspectual class of the input clause. The markers ....

M. Nagao. A framework of a mechanical translation between japanese and english by analogy principle. In A. Elithorn and R. Banerji, editors, Artificial and Human Intelligence. North-Hollan, 1984.

....selects a tutorial in the Tutoring Environment (see 3.4) the corresponding data are copied from this part of the database. The second part of the database comprises bilingual examples of English and Malay. The reason for this is that, the Query Translator (see 3. 5) uses an example based approach [5], to avoid the need for complex linguistic analysis and complex computations. However, this approach is highly dependent on a large number of examples in the database. At this stage of system development, a database of examples has been compiled for a demo version of the prototype. For now, the ....

M. Nagao. A framework of a mechanical translation between japanese and english by analogy principle. In Artificial and Human Intelligence. Amsterdam, 1984.

....the last few years this situation has changed considerably. In particular in Japan there have been extensive efforts regarding the automatic translation from Japanese into English and vice versa. One of the most prominent research directions in Japan has been example based machine translation [9, 10, 11], which relies on massive bilingual corpora to build a knowledge base of translation examples. New sentences are then translated by finding the most similar example. Unfortunately, this promising approach can only be applied successfully to language pairs for which enough bilingual data is ....

M. A. Nagao, "A Framework of a Mechanical Translation between Japanese and English by Analogy Principle", A. Elithorn and R. Banerji (eds), Artificial and Human Intelligence, North-Holland, Amsterdam, 1984.

....The differences in the minimal match sequence are replaced to create a SLGG of those strings. We also introduce a learning heuristic based on SLGGs of strings to be used in example based machine translation. 1 Introduction Example Based Machine Translation (EBMT) originally proposed by Nagao [15], is one of the main approaches of corpus based machine translation in which the required knowledge resources are automatically acquired from corpora. The main idea behind EBMT is that a given input sentence in the source language is compared with the example translations in the given bilingual ....

Nagao, M. A., Framework of a Mechanical Translation between Japanese and English by Analogy Principle, in: Artificial and Human Intelligence, A. Elithorn and R Banerji (eds.), NATO Publications, 1984.

....machine translation (MT) approaches have achieved considerable progress over the past 50 years, two new MT techniques have in recent years emerged, taking advantage of the rapidly improving computing power. The two emergent techniques are Statistical MT (Brown et al. 1990) and Example Based MT (Nagao 1984, Sato 1991) Both approaches require a huge quantity of translation examples from large existing translation corpora. Automatic alignment is an answer to such a need. Several methods of automatic alignment have already been reported in the literature. Among them, we find basically three ....

Nagao, Makoto: 1984, `A framework of a mechanical translation between Japanese and English by analogy principle', in Artificial and Human Intelligence, eds. Elithorn and R. Baner, Elevier Science Publishers, B.V.

....retrieved by the system. These structural correspondences will be used as raw translation patterns in a corpus based translation system. 1 Introduction So far, a number of methodologies and systems for machine translation using large corpora exist. They include example based approaches [7, 8, 9, 12], pattern based approaches [10, 11, 14] and statistical approaches. For instance, example based approaches use a large set of translation patterns each of which is a pair of parsed structures of a source language fragment and its target language translation fragment. Figure 1 shows an example of ....

Nagao, M., "A Framework of a Mechanical Translation between Japanese and English by Analogy Principle," Elithorn, A. and Banerji, R. (eds.) : Artificial and Human Intelligence, NATO 1984.

....are understood often rely on a full semantic and pragmatic analysis of the text, which is rarely available. 1. 2 Example based Translation To overcome the problems of dictionaries and rules in the rule based translation method, a method of translation by the principle of analogy has been proposed (Nagao 1984). This is done by collecting aligned translated example sentences and translating the input sentence by imitating the translation of a sentence that resembles it. This has its merits in that, as long as there is a translated example, a well structured translation will be generated. There is no ....

Nagao, Makoto: 1984, `A framework of mechanical translation between Japanese and English by analogy principle', in Elithorn & Banerji, eds., Artificial and Human Intelligence, Elsevier, pp.

....to other language pairs by making extensive reuse of the already available linguistic knowledge. At the sentence level we follow the transfer based translation paradigm by applying the translation rules learnt from the examples provided by the user. Unlike pure example based machine translation [13, 18, 21] we do not compare new sentences with all previous translation examples but we abstract from the examples by generating translation rules. However, we still keep a record of all relevant examples to verify the effect of a new translation rule on other examples. This is necessary to guarantee the ....

NAGAO, M., A Framework of a Mechanical Translation between Japanese and English by Analogy Principle, in: A. Elithorn and R. Banerji (eds.), Artificial and Human Intelligence, North-Holland, Amsterdam 1984.

....such as the lexicon are acquired automatically from a corpus [10] The technique presented here aims at acquiring all required knowledge except morphological rules for the machine translation task from sentence level aligned bilingual text corpora only. EBMT, originally proposed by Nagao [13], is one of the main approaches of corpus based machine translation. The main idea behind EBMT is that a given input sentence in the source language is compared with the example translations in the given bilingual parallel text to find closest matching examples so that these examples can be used ....

Nagao, M. A., Framework of a Mechanical Translation between Japanese and English by Analogy Principle, in: Artificial and Human Intelligence, A. Elithorn and R Banerji (eds.), NATO Publications, 1984.

....corpus) to derive the translation of an input. While statistical MT techniques use statistical metrics to choose the most probable structures in the target language, EBMT techniques employ pattern matching techniques to translate subparts of the given input. EBMT, originally proposed by Nagao [17], is one of the main approaches of corpus based machine translation. The main idea behind EBMT is that a given input sentence in the source language is compared with the example translations in the given bilingual parallel text to find the closest matching examples so that these examples can be ....

Nagao, M. A., Framework of a Mechanical Translation between Japanese and English by Analogy Principle, in: Artificial and Human Intelligence, A. Elithorn and R Banerji (eds.), NATO Publications, 1984.

....a new corpus. We experimented with metonymy interpretation and obtained a precision rate of 66 when using this method. Key words: metonymy, ellipsis, example based method, corpus 1. INTRODUCTION This paper describes a new Japanese metonymy interpretation method using the examplebased method (Nagao 1984; Murata and Nagao 1997; Murata et al. 1999b; Murata et al. 1999a) Metonymy is a metaphorical expression in which the name of something is substituted for another thing associated with the thing named. For example, in the Japanese sentence of boku ga torusutoi wo yomu (I read Tolstoi) the word ....

Nagao, M. 1984. A Framework of a Mechanical Translation between Japanese and English by Analogy Principle. Artificial and Human Intelligence, pages 173--180.

....for some of the approaches which are not discusses in this paper. These approaches include pure statistically based approaches (Brown et al. 5] the Shake and Bake approach (Beaven [4] knowledge based approaches (Goodman Nirenburg [8] Goodman [7] and example based approaches (Nagao [19], Somers [26] 2 Linguistic Approaches to MT Three main linguistic approaches to machine translation are surveyed, namely direct, transfer, and interlingua. The objective of this section is to give an overview of what these linguistic methods are, how they work, and what their advantages and ....

M. Nagao. A framework of a mechanical translation between Japanese and English by analogy principle. In A. Elithorn and R. Banerji, editors, Artificial and Human Intelligence, p.173-180, North Holland, Amsterdam, 1984.

....to provide sample translations. A statistical model is trained on the samples, and it is used to translate new sentences. Most of this research can be classified as either statistical machine translation (SMT) or examplebased machine translation (EBMT) it includes work such as (Brown et al. 1993; Nagao 1984; Wu 1997; Alshawi, Buchsbaum, Xia 1997; Melamed 2000; Och, Tillmann, Ney 1999) In this area, it is a truism that there s no data like more data. If a program sees a particular word or phrase one thousand times during training, it is more likely to learn a correct translation pattern than ....

Nagao, M. 1984. A framework of a mechanical translation between Japanese and English by analogy principle. In Elithorn, A., and Bernerji, R., eds., Artificial and Human Intelligence. North-Holland.

....of clustering to make retrieval of the best matching example from the database more efficient. Results on a large number of test cases from the CELEX database are presented. 1 INTRODUCTION EBMT is based on the idea of performing translation by imitating translation examples of similar sentences [Nagao 84] In this type of translation system, a large amount of bi multi lingual translation examples has been stored in a textual database and input expressions are rendered in the target language by retrieving from the database that example which is most similar to the input. There are three key ....

Nagao M., (1984). "A framework of a mechanical translation between Japanese and English by analogy principle". Artificial and Human Intelligence, ed. Elithorn A. and Banerji R., NorthHolland, pp 173-180.

....Translation is a new paradigm of machine translation, which emphasizes the significance of translation examples. The basic idea of Example Based Translation is very simple: it is to translate a sentence by using translation examples of similar sentences. After the first proposal by Nagao [1], researchers including the author have studied the implementation of the idea [2] MBT3 [3] is the latest model in a series of MBT by the author [5] MBT3 is designed for translation of technical terms, a difficult part of translation. A technical term is a noun phrase with several words in most ....

Nagao, M., A Framework of a Mechanical Translation between Japanese and English by Analogy Principle, in Elithorn, A. and Barnerji, R. (Eds.), Artificial and Human Intelligence, North-- Holland, pp173--180, 1984.

....languages for which little data is available. 1 Introduction The example based machine translation engine used in the Pangloss and DIPLOMAT projects (Brown 1996) has, until now, been purely lexical. Unlike other EBMT systems which include parsers and perform similarity matching on parse trees (Nagao 1984; Sato 1992) find the most similar complete sentences and modify their stored translations to generate a translation (Veale Way 1997) or produce an optimal partition of the input (Maruyama Watanabe 1992; Nirenburg et al. 1993) the DIPLOMAT EBMT engine performs overlapping partial exact ....

Nagao, M.: 1984, `A Framework of a Mechanical Translation between Japanese and English by Analogy Principle', in A. Elithorn & R. Banerji (eds), eds., Artificial and Human Intelligence, NATO Publications.

....(Kume et al. 1990) Shirai et al. 1990) However, the complexity of Japanese tense aspect modality expressions makes it very difficult to formulate detailed rules. We therefore tried to translate Japanese tense aspect modality expressions using the example based method, which was developed by Nagao (Nagao 1984). We prepared bilingual corpora containing pairs of Japanese and English sentences and tried translating tense aspect modality expressions by using the tense aspect modality expression of the English sentence corresponding to the most similar Japanese sentence. The example based method developed ....

....We prepared bilingual corpora containing pairs of Japanese and English sentences and tried translating tense aspect modality expressions by using the tense aspect modality expression of the English sentence corresponding to the most similar Japanese sentence. The example based method developed by Nagao in 1984 is effective but has rarely been used since it was used by Sumita et al. Sumita et al. 1990) in the translation of the Japanese particle no 1 . The method we describe here is the first application of the example based method to tense aspect modality translation. It is based on a very simple ....

Nagao, Makoto: 1984, `A Framework of a Mechanical Translation between Japanese and English by Analogy Principle', Artificial and Human Intelligence, pp. 173--180.

....lexicalised tree adjoining grammars can. Bahl et al. s system can also include (in theory) semantic and pragmatic features of word context, as can Bod s data oriented parsing scheme. Both these systems share some structural similarities with representations in example based translation [97, 70]. Fisher and Riloff [50] use the t statistic as a measure of co occurrence likelihood between two items. It too is calculated from corpus frequency information and can indicate strong correlations between items. This statistic can measure collocational differences, whereas mutual information ....

Makoto Nagao. A framework of a mechanical translation between Japanese and English by analogy principle. In A. Elithorn and R. Barnerji, editors, Artificial and Human Intelligence, pages 173 -- 180. North-Holland, 1984.

....Report Series Department of Computer Science TMI 95 WU 2 1 Introduction Phrasal translation examples at the subsentential level are an essential resource for many MT and MAT architectures. This requirement is becoming increasingly direct for the example based machine translation paradigm (Nagao 1984), whose translation flexibility is strongly restricted if the examples are only at the sentential level. It can now be assumed that a parallel bilingual corpus may be aligned to the sentence level with reasonable accuracy (Kay Roscheisen 1988; Catizone et al. 1989; Gale Church 1991; Brown et ....

NAGAO, MAKOTO. 1984. A framework of a mechanical translation between japanese and english by analogy principle. In Artificial and human intelligence: Edited review papers presented at the International NATO Symposium on Artificial and Human Intelligence, ed. by Alick Elithorn & Ranan Banerji, 173--180. Amsterdam: North-Holland.

.... the most probable words in the target language, EBMT uses pattern matching techniques to translate subparts of the given input [1] Also, Siskind has dealt with the question of what matches what in the examples in some detail for the case of lexical matching [12] EBMT has been proposed by Nagao [7] as Translation by Analogy which is in parallel with memory based reasoning [13] case based reasoning [9] and derivational analogy [2] The example based translation relies on the use of past translation examples to derive a translation for a given input. The input and the example translations ....

Nagao, M. A.: Framework of a Mechanical Translation between Japanese and English by Analogy Principle (1985).

....for some of the approaches which are not discusses in this paper. These approaches include pure statistically based approaches (Brown et al. 6] the Shake and Bake approach (Beaven [5] knowledge based approaches (Goodman Nirenburg [9] Goodman [8] and example based approaches (Nagao [22], Somers [28] The reader is referred to these references for further details. 2 Linguistic Approaches to MT This section surveys three main linguistic approaches to machine translation, namely direct, transfer, and interlingua. The objective of this section is to give an overview of what ....

M. Nagao. A framework of a mechanical translation between Japanese and English by analogy principle. In A. Elithorn and R. Banerji, editors, Artificial and Human Intelligence, p.173-180, North Holland, Amsterdam, 1984.

....to carry out translation based on complex cooccurrence and distribution probability calculations over very large aligned bilingual text corpora; and ffl a more modest approach, called example based MT, which is the topic of this paper. Example Based MT The basic idea of EBMT is simple (cf. Nagao, 1984): given an input passage S in a source language and a bilingual text archive, where text passages S 0 in the source language are stored, aligned with their translations, T 0 , into a target language, S is compared with the source language side of the archive. The closest match for passage ....

Nagao, M. 1984. A framework of a mechanical translation between Japanese and English by analogy principle. In: A. Elithorn and R. Banerji (eds.) Artificial and Human Intelligence.

....translation (called DICT or GLOSS below) and corpus based term translation (called EBT below, for Example Based Term translation) are further described. All three MT based methods used variations of the Pangloss Example Based Machine Translation engine (PanEBMT) 3] In general, EBMT systems [3, 18] use a large corpus of example pairs of previously translated sentences in order to find close matches and translations of words and phrases in context. The PanEBMT parallel corpus was derived primarily from the Spanish and English portions of the UN Multilingual Corpus [14] with an admixture of ....

M. Nagao. A Framework of a Mechanical Translation between Japanese and English by Analogy Principle. In A. Elithorn and R. Banerji (eds), editors, Artificial and Human Intelligence. NATO Publications, 1984.

....in memory without any change in the representation. The characteristic examples stored in the memory are called exemplars. The basic idea in exemplar based learning is to use past experiences or cases to understand, plan, or learn from novel situations [4, 6, 10] EBMT has been proposed by Nagao [8] as Translation by Analogy which is in parallel with memory based reasoning [14] case based reasoning [11] and derivational analogy [2] Example based translation relies on the use of past translation examples 1 This research has been supported in part by NATO Science for Stability Program ....

Nagao, M. A.: Framework of a Mechanical Translation between Japanese and English by Analogy Principle (1985).

....just to devise a principled way to predict which verb a given item will take. However, the information needed to learn these distinctions is available in the huge amount of already translated text available from various sources worldwide. The now mainstream Example Based Translation paradigm (e.g. [53, 62]) relies on exactly this sort of information to produce translations with a reduced but still significant amount of human engineering. By encoding a tremendous database of sentences or fragments together with their translations, a system can select the most appropriate (for example, closest to the ....

....capture the so called rules of grammar more directly) they require that the inevitable exceptions to every rule of grammar be coded as well. Like any large rule based system, the interactions among transfer rules are not always obvious and amenable to correction. Example based translation Nagao[53] outlined what would become a major force in machine translation with his proposal of what would become known as example based translation. Rather then generating explicit rules, translation systems would store a huge collection of translation examples that would provide coverage in context for ....

Makoto Nagao. A framework of a mechanical translation between Japanese and English by analogy principle. In A. Elithorn and R. Barnerji, editors, Artificial and Human Intelligence, pages 173--80. North-Holland, 1984.

....Pangloss Example Based Machine Translation engine (PanEBMT) Brown, 1996 ] however only corpus based term translation (called EBT below, for Example Based Term translation) is further described since it produced better results. 2. 1 PanEBMT Translations In general, EBMT systems [ Brown, 1996; Nagao, 1984 ] use a large corpus of example pairs of previously translated sentences, in order to find close matches and translations of words and phrases in context. The PanEBMT parallel corpus was derived primarily from the Spanish and English portions of the UN Multilingual Corpus [ Graff and Finch, 1994 ] ....

M. Nagao. A Framework of a Mechanical Translation between Japanese and English by Analogy Principle. In A. Elithorn and R. Banerji (eds), editors, Artificial and Human Intelligence. NATO Publications, 1984.

....contrast, lexical transfer is resolved using bilingual constraints. The qualitative and quantitative knowledge is applied interleavingly and cooperatively, so that the advantages of both approaches are kept. 1. Introduction Many different approaches (Bennett and Slocum, 1985; Brown et al. 1990; Nagao, 1984; Mitamura et al. 1991; Baker et al. 1994) to machine translation system design have been proposed in literature. Traditional rule based machine translation system (Bennett and Slocum, 1985) involves many human costs in formulating rules. That easily introduces inconsistencies, and it is too ....

....processing partial and ill formed sentences. However, the computation time in a statistics based system increases potentially with the length of sentences. In additional, the parameters strongly depend on the training corpus, i.e. it is domain dependent. The performance of an example based system (Nagao, 1984) depends on the quality of collected examples and the similarity measure on examples and input sentences. When the matched units are subsentential structures (phrase structures) the performance of such a system is better than that of a word level system. As for knowledge based system (Mitamura et ....

Nagao, M. (1984), "A Framework of Mechanical Translation between Japanese and English by Analogy Principle," Artificial and Human Intelligence (Elithorn, A. Eds.), 1984, pp.

....retrieval through the tree structure of a thesaurus, 3) binary search along subsumption ordering of retrieval queries. Example retrieval time drastically decreases with the method. 1 Introduction Since a model of machine translation (MT) called Translation by Analogy was first proposed in Nagao (1984), much work has been undertaken in examplebased NLP (e.g. Sato and Nagao (1990) and Kurohashi and Nagao (1993) The basic idea of examplebased approach to NLP is to accomplish some task in NLP by imitating a similar previous example, instead of using rules written by human writers. Major ....

Nagao, M. (1984). A framework of a mechanical translation between Japanese and English by analogy principle, in A. Elithorn and R. Banerji (eds), Artificial and Human Intelligence, Elsevier Science Publishers, B.V.

No context found.

M. Nagao. 1984. A framework of a mechanical translation between Japanese and English by analogy principle. In: A.Elithorn and R.Banerji (eds.) Artificial and human intelligence (Amsterdam: North-Holland), 173-180.

No context found.

Makoto Nagao. 1984. A Framework of a Mechanical Translation between Japanese and English by Analogy Principle. In Alick Elithorn and Ranan Banerji, editors, Artificial and Human Intelligence, pages 173-- 180, Lyon, France, October, 1981. Elsevier Science Publishers B.V. Proceedings of the International NATO Symposium.

No context found.

M. Nagao, \A framework of mechanical translation between Japanese and English by analogy," in Arti cial and Human Intelligence (A. Elithorn and R. Banerji, eds.), pp. 173-180, North Holland, Amsterdam, 1984.

No context found.

M. Nagao. A Framework of a Mechanical Translation between Japanese and English by Analogy Principle. Nato Publications, 1984.

No context found.

M. Nagao. A framework of a mechanical translation between Japanese and English by analogy principle. In A. Elithorn and R. Banerji, editors, Artificial and Human Intelligence, pages 173--180. North-Holland, Amsterdam, 1984.

No context found.

Nagao, M., A Framework of a Mechanical Translation Between Japanese and English by Analogy Principle, In Artificial and Human Intelligence, Amsterdam, 1984.

No context found.

Makoto Nagao. 1984. A Framework of a Mechanical Translation between Japanese and English by Analogy Principle. Artificial and Human Intelligence, pages 173--180.

No context found.

M. Nagao. 1984. A framework of a mechanical translation between Japanese and English by analogy principle. In: A. Elithorn and R. Banerji (eds.) Artificial and Human Intelligence.

No context found.

Nagao, M. (1984) . "A Framework of a Mechanical Translation between Japanese and English by Analogy Principle," Artificial and Human Intelligence, Elithorn, A. and Banerji, R. (eds.), Elsevier Science Publishers, B. V. 1984.

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC