File Information
File: 05-lr/acl_arc_1_sum/cleansed_text/xml_by_section/metho/98/p98-2141_metho.xml
Size: 20,022 bytes
Last Modified: 2025-10-06 14:15:00
<?xml version="1.0" standalone="yes"?> <Paper uid="P98-2141"> <Title>Simultaneous Interpretation Utilizing Example-based Incremental Transfer</Title> <Section position="2" start_page="855" end_page="857" type="metho"> <SectionTitle> 1 Incremental Translation Using Transfer-Driven Machine Translation </SectionTitle> <Paragraph position="0"/> <Section position="1" start_page="855" end_page="855" type="sub_section"> <SectionTitle> 1.1 Constituent Boundary Pattern </SectionTitle> <Paragraph position="0"> In TDMT, translation is performed by applying stored empirical transfer knowledge, which describes the correspondence between source language expressions and target language expressions at various linguistic levels. The source and target expressions from the transfer knowledge in TDMT are expressed as CB-Patterns, which represent meaningful units for linguistic structure and transfer. The efficient application of transfer knowledge source components to an input string plays a key role in our basic incremental translation scheme. A pattern is defined as a sequence that consists of variables and constituent boundaries such as surface functional words.</Paragraph> <Paragraph position="1"> The transfer knowledge is compiled from actual translation examples in every source pattern.</Paragraph> </Section> <Section position="2" start_page="855" end_page="856" type="sub_section"> <SectionTitle> 1.2 Incremental Pattern Application </SectionTitle> <Paragraph position="0"> The incremental application of CB-Pattems is based on the idea of incremental chart parsing (Furuse, 1996) (Amtrup, 1995) with notions of linguistic levels.</Paragraph> <Paragraph position="1"> The procedure for the application of CB-Patterns is as follows: (a) Determination of possible pattern applications. (b) Translation candidate determination and structural disambiguation of patterns by semantic distance calculation.</Paragraph> <Paragraph position="2"> Our scheme determines the best translation and structure parallel with an input sequence and can restrain the number of competing structures (possible translation candidates) at the possible utterance point in the input by performing (a) in parallel with (b), thus reducing the translation costs in time. The structure selected in (b) has its result transferred with head word-information using semantic distance calculations when combined incrementally with other structures. The output sentence is generated as a translation result from the structure for the whole input, which is composed of best-first substructures.</Paragraph> <Paragraph position="3"> In order to limit the combinations of patterns and control the appropriate timing of each partial utterance during pattern application, we distinguish pattern levels, and specify the linguistic sublevel permitted for use in the assigned variables for each linguistic level. This is because if any combinations of patterns are permitted, it is obvious that the possibility of combinations are easily exploded. Table 1 shows examples of the relationship between linguistic levels. Every CB-pattern is categorised as one of the linguistic levels, and a variable on a given level is instantiated by a string on the linguistic levels in the second column of Table 1.</Paragraph> <Paragraph position="4"> For instance, in the noun phrase &quot;X of Y&quot;, the variables X and Y cannot be instantiated by a simple sentence pattern, but can be instatiated by NP such as a noun phrase pattern or a compound noun pattern.</Paragraph> <Paragraph position="5"> Moreover, these levels give a guideline to the timing of utterance production (i.e. the timing of when an utterance is said). For example, each simple sentence level pattern has utterance markers (Table 2, where '/' indicates the utterance markers) for possible insertion of an utterance during left-to-right application of the pattern. Thus, redundant or incomplete partial matchings can be eliminated and an appropriate trigger of utterance can be obtained.</Paragraph> <Paragraph position="6"> (Furuse, 1996) provides further details of the algorithm for incremental CB-Parsing.</Paragraph> <Paragraph position="7"> Table 1 Possible linguistic sublevels in variables</Paragraph> <Paragraph position="9"/> </Section> <Section position="3" start_page="856" end_page="857" type="sub_section"> <SectionTitle> 1.3 Disambiguation of Translation Candidate </SectionTitle> <Paragraph position="0"> The CB-pattern &quot;X no Y&quot; with the particle &quot;no&quot; is a frequently used expression in Japanese. We can observe the following Japanese-to-English transfer knowledge about &quot;X no Y'&quot; from such translation examples as the source-target pairs of : &quot;hoteru no jasho'&quot; --~ &quot;the address of the hotel&quot;, &quot;eigo no paNfuretto'&quot; ~ &quot;the pamphlet written in English&quot;, etc.</Paragraph> <Paragraph position="2"> Within this pattern, X' is the target word corresponding to X, and a corresponding English word is written below each Japanese word. For example, &quot;hoteru'&quot; means 'hotel', and &quot;jasho &quot; means 'address'.</Paragraph> <Paragraph position="3"> This transfer knowledge expression indicates that the Japanese pattern &quot;X no Y&quot; corresponds to many possible English expressions. (hoteru, jasho) are sample bindings for &quot;X no Y&quot;, where X = hoteru, and Y = jasho.</Paragraph> <Paragraph position="4"> TDMT makes the most of an example-based framework, which produces an output sentence by mimicking the closest translation example to an input sentence. The semantic distance from the input is calculated for all examples. Then the example closest to the input is chosen, and the target expression of that example is extracted.</Paragraph> <Paragraph position="5"> Suppose that the input is &quot;nihoNgo no paNfuretto&quot;, where nihoNgo means 'Japanese', and the input is closest to (eigo, paNfuretto); &quot;the pamphlet written in Japanese&quot; can be gained by choosing Y' written in X' as the best target expression.</Paragraph> <Paragraph position="6"> Furthermore, ambiguity in the combination of patterns, which have not been constrained by the linguistic levels, is also dissolved incrementally by using the total sum of the semantic distances of patterns contained (Furuse, 1996).</Paragraph> <Paragraph position="7"> The distance between an input and a translation example is measured based on the semantic distance between the words contained, and the semantic distance between words is calculated in terms of a thesaurus hierarchy. (Sumita, 1991) provides further details of the semantic distance caluculation.</Paragraph> <Paragraph position="8"> 2 Exploitation of a Simultaneous In practical simultaneous interpretation, human translators generally use strong sentence planning such as transformation between the active and the passive voice, transformation from a lengthy interrogative sentence to a tag question, and topicalization transformation. Moreover, the input is produced and modified in a step-by-step manner, so that it can be temporarily incomplete - although as a whole sentence it may become sufficient.</Paragraph> <Paragraph position="9"> Thus, the consistency of translations has to be adjusted appropriately when a contradiction occurs between a previously uttered part of the translation and the part currently being translated.</Paragraph> <Paragraph position="10"> As a consequence of under specification, simultaneous interpretation is essentially based on working with empirical knowledge - e.g.</Paragraph> <Paragraph position="11"> simultaneous interpreters' translation examples. In this section, we first describe the kinds of examples that are required to achieve simultaneous interpretation using some sample sentences.</Paragraph> </Section> <Section position="4" start_page="857" end_page="857" type="sub_section"> <SectionTitle> 2.1 Empirical Knowledge </SectionTitle> <Paragraph position="0"> * Transformation to a tag question Let us consider the following Japanese utterance: (El) Nani-mo moNdai-wa ari-maseN -<pause>de-sh~-ka. (what problem exist -<pause>- is there) 1 In Japanese, an interrogative is specified at the end of the sentence, while in English, it is generally specified in front of the sentence. Thus, although a translation of the whole sentence of (El) is &quot;Is everything all right', in some cases, &quot;Everything is all right' could be uttered after the pause in the incremental framework. In this case, the meaning of the previously uttered part is no longer consistent with the current translation.</Paragraph> <Paragraph position="1"> However, even in this case, translation can be continued transforming to a tag question as (El)' by using a peculiar translation example \[TEll without interruption by semantic inconsistency and the insertion of a restatement.</Paragraph> <Paragraph position="2"> In Japanese, negation is also specified at the end of the sentence while in English it has to be specified in front of the finite verb. In addition, an expression &quot;X wa gozai-masu&quot; in (E2) has possible translations as &quot;'we have X'&quot; or &quot;X' is available&quot;. Thus, although the whole translation should ideally read as &quot;We have twin rooms, but none are available today&quot;, &quot;A twin room is available&quot; might be selected as a part of the translation in some cases. Although one solution could be to restate previously uttered phrases such I In this paper, sample Japanese is Romanized in italic based on the Hepburn system with the corresponding English words following in parentheses.</Paragraph> <Paragraph position="3"> as: &quot;no, sorry, we do have twin rooms, but none ..... &quot;, such restatements should not be used frequently. This is because the restatements tend to break in general, coherency of human interaction However, in this case, translation can be continued as (E2)' by using a peculiar translation example \[TE2\], with no restatement.</Paragraph> <Paragraph position="4"> \[TE2\] (X ga, Y) ---) (X' usually, but Y') (E2)' A twin room is available usually, but we do not have any vacancies today. ({\[TE2\]: X'= 'A twin room is available', Y'='we do not have any vacancies today' }) * Failure of prediction In simultaneous interpretation, elements are usually uttered before the input consumption has been finished. Thus, because of the uncertainty in assumptions, a system with this facility must be able to adjust the whole content of the translation when it is realized that the assumption is incorrect from information given later.</Paragraph> <Paragraph position="5"> Consider the following English utterance: (E3) That restaurant is open -<pause>- as only as in the evening.</Paragraph> <Paragraph position="6"> In the case of the part of the translation already uttered, &quot; sono-resutoraN-wa @uN-shite-I-masu&quot;, it should have been inserted &quot;yoru nomi&quot; in front of the phrase &quot;@uN-shite-l-masu&quot;, when the whole sentence is translated.</Paragraph> <Paragraph position="7"> However the translation can be continued as it is as in (E3)' by using a peculiar translation example</Paragraph> <Paragraph position="9"> As the above example shows, simultaneous interpretation as skilled as that performed by a human interpreter is achievable by exploiting peculiar translation examples - i.e. simultaneous interpretation examples (or SI-examples, in short).</Paragraph> <Paragraph position="10"> In the next section, we propose an algorithm to handle these kinds of SI-example with the best-first example-based incremental MT mechanism.</Paragraph> </Section> </Section> <Section position="3" start_page="857" end_page="859" type="metho"> <SectionTitle> 3 Simultaneous Interpretation Algorithm </SectionTitle> <Paragraph position="0"> Although the main characteristic of example-based translation is the use of the most similar examples as the main knowledge source for translation, the exploitation of SI-examples is drawn from the following consideration : * A translation should use an example consistent with previously uttered information Thus, the key translation process with exploiting SI-examples consists of the following stages: (1) Checking the contextual consistency between previously uttered phrases 2 and the phrase to be uttered next.</Paragraph> <Paragraph position="1"> (2) Retrieving the most plausible example according to both the contextual sequence and similarity.</Paragraph> <Paragraph position="2"> (3) Re-translating the phrase to be uttered next by using the example retrieved in (2) The algorithm is described as follows. In the algorithm, the input phrase to be considered as a combination of structures shown in Figure 1 to facilitate understanding of the algorithm. For example, in the case of (E3), STj indicates &quot;'The restaurant is open&quot;, ST2 indicates &quot;open as only as in the evening&quot;, and STy.2 indicates the whole phrase. In addition, trans(S%) returns word sequence indicating translation of S%., trans(STi, E) also returns word sequence indicating the translation of S% using example E, and i indicates the current processing part. Since the algorithm for the exploitation of SI-examples is applied only if a previous translated phrase exists, the proposed algorithm is executed in the case of i>=2.</Paragraph> <Paragraph position="3"> Algorithm: Start.</Paragraph> <Paragraph position="4"> 1. Retrieve the similar examples of ST~ from the total example database (normal + SI-examples) and assign the list to the {SE} with the appropriate semantic distance.</Paragraph> <Paragraph position="5"> 2. Produce trans(ST~, E), where E indicates the most similar example listed in {SE}.</Paragraph> <Paragraph position="6"> 3. Remove the example E from {SE}.</Paragraph> <Paragraph position="7"> 4. If trans(STi.,.~., E) == trans(STH) +3 trans(ST~, E) 4, 2 In this paper, we only state the context within a sentence and do not refer to contexts between dialogs. 3 Indicating sequencial appending operation, which includes removal operation of the common sub-sequence among the last of the first item and the first of the second item. For example, word sequences &quot;A B&quot; + word sequences &quot;B C&quot; indicates &quot;A B C&quot;. 4 i.e. trans(STi.j) and trans(STi) are contextually continuous. In this paper, we define contextually continuous from the view point of sequences of concrete words (phrases) contained, in terms of combination with an example-based framework. J j i trans(ST,): &quot;i i i Sono'resutoraN'waldPuN'shlte I.masu i iF,,- i Ou trans(ST2): i @mV-slffta I.mssu, k.g.a..,...yo..rU.....no.m..i.:.de.s_~ Figure 1 Notation of Substructures then, output the difference between trans(STi, E) and trans(ST~. 0, then goto End.</Paragraph> <Paragraph position="8"> 5. Goto 2.</Paragraph> <Paragraph position="9"> End.</Paragraph> <Paragraph position="10"> In the majority of conventional example-based frameworks, only a semantic similarity is considered in retrieving the examples to be applied. In our scheme, on the other hand, not only semantic similarity but also contextual consistency with the previous translation is considered. In other words, the key notion of the scheme is its mechanism for selecting appropriate examples. Hence, as the above algorithm shows, exploitation of SI-examples can be combined smoothly with the conventional example-based framework.</Paragraph> <Paragraph position="11"> Let us explain the algorithm in terms of sentence (E3) as an example. First, assuming that trans(STl) = &quot;Sono-resutoraN-wa &puN-shite Imasu&quot; (the-restaurant open), the most similar example of ST~ is normally: \[TE4\] (X as only as Y) ---) (Y' nomi X' I-masu) Thus, trans(ST2, TE4) can be &quot;yoru nomi &puN-shite l-masu'&quot; (evening only open) and as the phrase &quot;yoru nomi ...&quot; is, in this case, not contextually continuous, and the next example should be extracted from the similar example list {SE}. Then, the example is \[TE3\], since trans(ST2, TE3) -- &quot;&ouN-shite l-masu, ga, yoru nomi-desu&quot;, in terms of the contextual order of the words, this translation can be continuous. Thus, the difference between trans(ST,) and the trans(ST2, TE3), &quot;ga, yoru nomi-desu&quot; can be obtained as the next utterance.</Paragraph> </Section> <Section position="4" start_page="859" end_page="859" type="metho"> <SectionTitle> 4 Preliminary Experiments </SectionTitle> <Paragraph position="0"> We conducted a preliminary experiment with respect to (a) the quality of example-based translation in relation to IUs (i.e., meaningful units), and (b) the quality and speed of incremental parsing (CB-Parsing), to confirm the feasibility of our proposed scheme.</Paragraph> <Paragraph position="1"> In the evaluation of (a), we conducted a jackknife experiment to measure the average success rate of translation for the most frequently used expressions (i.e. the most ambiguous) in Japanese, &quot;X no Y'&quot; and &quot;X wo Y&quot;. We prepared 774 and 689 examples for the expressions respectively, and conducted the experiment in increments of 100 examples (Furuse, 1994a). The examples were extracted by random sampling. We then evaluated the 10 translations of corresponding expressions in the dialog database for each case.</Paragraph> <Paragraph position="2"> Figure 2 shows the average rate of the evaluation for 10 translations.</Paragraph> <Paragraph position="3"> Although the translation quality of each unit depended on the type of expression, the graph shows that, in general, the more examples the system has, the better the quality 5.</Paragraph> <Paragraph position="4"> Conditions of our experiment and evaluation for (b) are that the number of CB-patterns for Japanese-English translation and English-Japanese translation are 777 and 1241, respectively, and the number of total examples are 10000 and 8000, respectively. In the evaluation, we set the system to retain only one substructure in the semantic distance calculation in order to confirm the feasibility of deterministic processing at each incremental step.</Paragraph> <Paragraph position="5"> CB-Parsing for 69-77 unseen dialogs (of 1,000 different unseen sentences) were manually evaluated by assigning a grade indicating success or failure. All of the parsing times include accessing time for an example database (i.e.</Paragraph> <Paragraph position="6"> corresponding to the whole transfer time) and were measured on a Sparc Station 10 workstation with 256 MB of memory.</Paragraph> <Paragraph position="7"> Table 3 shows the experimental results. For CB-Parsing accuracy, a success rate of approximately 76 % was achieved for both translations, rates that are fairly high for spoken-language parsing.</Paragraph> <Paragraph position="8"> 5However, we also have to ascertain the practical satiation limit, or how much the transfer knowledge can be expanded, as a future work.</Paragraph> <Paragraph position="9"> The main problem in the parsing procedure involved an insufficient number of examples for the CB-Pattem. However, as Figure 2 shows, an increase in the ratio with the number of examples could be observed with our framework. Thus, overall accuracy and acceptability should improve in proportion to an increase in transfer examples. Although the speed depends on the amount of knowledge and sentence length, the average time was less than 0.4 seconds, which is fairly rapid. Thus, our translation scheme can be seen as an efficient translation mechanism in achieving a practical simultaneous interpretation system.</Paragraph> </Section> class="xml-element"></Paper>