TWI644223B - Translation memory enhancement system - Google Patents

Abstract

The present invention provides a translation memory enhancement system including at least one electronic device and a backend server. The electronic device includes at least one file to be translated and a storage unit, and the file to be translated is composed of a plurality of sentences to be translated. The electronic device is connected to the backend server by means of a network connection, and the backend server comprises a historical translation memory, a comparison analysis unit, an integration unit, a replacement unit and a export unit. The historical translation memory includes a plurality of different first translation memories. The first translation memory includes an original text and a translation data. The original data is composed of multiple original clauses, and the translation data is composed of multiple translation clauses. And each of the original clauses corresponds to one of the translation clauses. The comparison analysis unit compares each of the to-be-translated clauses with all the original clauses of each original text to search for an original clause that approximates the original clause of the clause to be translated and its corresponding translation clause. Memory library.

Description

Translation memory enhancement system

The present invention provides a translation memory enhancement system, and more particularly to a translation memory enhancement system having alignment units.

Translation Memory is a commonly used database in the translation industry that can store translated content for later reuse. In the translation industry, computer-assisted translation software, such as Trados, is often used to generate translation memories. The translation memory records each sentence in the original text and its corresponding translation. When a translator's sentence is to be found in a similar sentence in the translation memory, the translator can use the translation provided by the translation memory or modify it to complete the translation. Therefore, the use of translation memories will help translators translate documents more efficiently and complete translations more quickly. However, the current use of translation memories is usually only for the same customer or the same product line, because the same customer or the same product line will have a higher similarity. The new translation cases come from different customers or product lines and often require a new translation of the new customer or new product line. Therefore, the initial translation will not have any translation memory for use, and it will not improve the translation efficiency of translators. Therefore, if a comparison technique can be provided, the new translation memory that can be applied to the new translation case from all the old translation memories will make the original translation of the original translation memory without a translation memory. The library is used to improve the efficiency of translation. Therefore, how to use the comparison technique to correctly and efficiently extract new translation memories that can be applied to new translation cases from the old translation memory is worthy of consideration in the field.

It is an object of the present invention to provide a translation memory enhancement system that enables new translation files for new customers or new product lines to have a suitable translation memory for improved translation efficiency. The present invention provides a translation memory enhancement system including at least one electronic device and a backend server. The electronic device includes at least one file to be translated and a storage unit, and the file to be translated is composed of a plurality of to-be-translated clauses. The electronic device is connected to the backend server by means of a network connection, the backend server includes a historical translation memory and a comparison analysis unit, and the historical translation memory includes a plurality of different first translation memories. The first translation memory includes an original text and a translation data. The original data is composed of a plurality of original clauses, the translation data is composed of multiple translation clauses, and each of the original clauses corresponds to one of the translation clauses. sentence. The comparison analysis unit compares each of the to-be-translated clauses with all the original clauses of each original text to search for the original clause of the clause to be translated and its corresponding translation clause to form an approximation. Translation memory. The translation memory enhancement system described above, wherein the backend server further includes an integration unit; in the approximate translation memory, there may be the same or similar translation memory records (sentences of the same or similar content), and the integration unit will The system presets or user-defined rules integrate these same or similar translation memory records to convert the approximate translation memory into a second translation memory. The translation memory enhancement system described above, wherein the backend server further comprises a replacement unit; when the second translation memory includes at least one brand text, the replacement unit replaces the brand text to make the second The translation memory is transformed into a third translation memory. In the above described translation memory enhancement system, the backend server further includes a export unit that exports the third translation memory to the storage unit. In the above described translation memory enhancement system, the electronic device further includes a display interface, and the original text clauses are organized into an analysis result according to the degree of approximating the to-be-translated clause, and the analysis result is transmitted to the comparison analysis unit. Display interface. The translation memory enhancement system described above, wherein the electronic device is a desktop computer, a notebook computer, a tablet computer or a smart phone. In order to make the above features and advantages of the present system more apparent, the preferred embodiments are described below, and are described in detail below with reference to the accompanying drawings.

Referring to FIG. 1 , FIG. 1 is a schematic diagram of a translation memory enhancement system 1 of the present embodiment. The translation memory enhancement system 1 includes at least one electronic device 12 and a back-end server 10 , and the electronic device 12 includes at least one The file to be translated 120, a display interface 121 and a storage unit 122. The electronic device 12 is, for example, a desktop computer, a notebook computer, a tablet computer or a smart phone, and the file to be translated 120 is the original manuscript to be translated by the user, that is, the translation case of the new client. The file to be translated 120 is composed of a plurality of to-be-translated clauses 120A, and the to-be-translated clause 120A and the to-be-translated clause 120A are separated by a rule formula of a sentence (the rule formula of the sentence is described in detail below). In addition, the electronic device 12 is connected to the backend server 10 by means of a network connection. The backend server 10 includes a historical translation memory 102, a comparison analysis unit 103, an integration unit 104, and a replacement unit 105. And a export unit 106. The historical translation memory 102 includes a plurality of different first translation memories 1020. The first translation memory 1020 includes an original text 1021 and a translation data 1022. The original data 1021 is composed of multiple original clauses 1021A. The original clause 1021A and the original clause 1021A are separated by a regular sentence formula, and the translation material 1022 is composed of a plurality of translation clauses 1022A. Each of the original text clauses 1021A will correspond to one of the translation clauses 1022A. For example, if the original clause 1021A (in Chinese) is "the film 1 NVDIA G-SYNC brings a smooth, no-screen tear experience", the corresponding translation clause 1022A (English example ) is "Video 1 See how NVIDIA G-SYNC brings smooth, tear-free viewing experience." In the above, the rule formula of the sentence is based on the rules of each language sentence specified by the computer-aided translation software Trados. For example, when the original material 1021 is in English, the following punctuation marks ".", "Tab", ":", " ?? or the paragraph mark "¶" as a criterion for separation between the original clause 1021A and the original clause 1021A. The comparison analysis unit 103 compares each of the to-be-translated clauses 120A with all the original clauses 1021A of each original document 1021 to search for the original clause 1021A of the to-be-translated clause 120A. In detail, the comparison analysis unit 103 compares the to-be-translated clause 120A and the original clause 1021A which are broken by the sentence-sentence rule formula. In addition, the comparison analysis unit extracts the translation clause 1022A corresponding to the original clause 1021A and the original clause 1021A of the to-be-translated clause 120A to form an approximate translation memory 1041. Then, when there is a repeated identical or similar content sentence in the approximate translation memory 1041, the integration unit 104 integrates the same or similar content sentences, and converts the approximate translation memory 1041 into a second translation memory 1042. . In detail, different original text data 1021 may have the same or similar original text clause 1021A, and therefore, when the original text clause 1021A corresponding to the to-be-translated clause 120A and the translation clause 1022A corresponding to the original clause 1021A are formed, a translation clause 1022A is formed. When an approximate translation memory 1041 is approximated, the integration unit 104 integrates the same or similar original clauses 1021A according to system presets or rules set by the user, thereby integrating into a more accurate (removing duplicate data) and being suitable for assisting the files to be translated. 120 translates the second translation memory 1042. Thereafter, if the second translation memory 1042 includes at least one brand text, the substitution unit 105 replaces the brand text to convert the second translation memory 1042 into a third translation memory 1051, and the third translation memory 1051 is stored in the replacement unit 105. In detail, since the original text data 1021 in the first translation memory 1020 is the data belonging to the old customer, in order to prevent the translator from appearing other brand characters after translating the to-be-translated file 120, the replacing unit 105 can replace the first The second translation of the brand text in the memory 1042. In this way, the errors of human translation can be reduced and the quality of translation can be improved. Finally, the export unit 106 exports the third translation memory 1051 to the storage unit 122. In this way, when the translator translates the translation file 120, the third translation memory 1051 of the storage unit 122 can be used to complete the translation. Therefore, the translation memory enhancement system 1 of the present embodiment can quickly establish a new translation memory of a new customer or a new product line to improve the efficiency of the translator. Please refer to FIG. 2 . FIG. 2 is a schematic diagram showing the analysis result 103R displayed on the display interface 121 . The first translation memory 1020 is organized into an analysis result 103R by the comparison analysis unit 103 according to the degree of approximating the clause to be translated, and the analysis result 103R is transmitted to the display interface 121 by the comparison analysis unit 103. In this way, the user of the electronic device 12 can know through the analysis result 103R which of the first translation memories 1020 is most similar to the file to be translated 120. For example, in FIG. 2, there are four first translation memories 1020 of A, B, C, and D, and the number of original clauses 1021A is also classified according to the degree of similarity. Among them, the first translation memory 1020 representing "C" has a total of 300 sentences in 97%~100% similarity, 610 sentences in the similarity degree of 85%~96%, and a total of 150 in the similarity degree of 51%~84%. Sentences, while sentences below 50% similarity are ineffective in the translation industry. Therefore, the first translation memory 1020 representing "C" has the most similar original clause 1021A in 51%~100%, far exceeding the other three (A, B, D) first translation memories 1020. The user can select a higher proportion of translation memories and set a similar scale to export the new translation memory. In summary, the translation memory enhancement system 1 of the present embodiment analyzes and compares the content of the historical translation memory 102 via the comparison analysis unit 103, and can actually generate a new file to help the new customer or the new product line to be translated 120. Translate memory to improve the efficiency of translator translation. Although the present invention has been disclosed in the above preferred embodiments, it is not intended to limit the invention, and any one of ordinary skill in the art can make some modifications and refinements without departing from the spirit and scope of the invention. Therefore, the scope of the invention is defined by the scope of the appended claims.

1‧‧‧ Translation Memory Enhancement System

10‧‧‧Backend server

102‧‧‧Historical translation memory

1020‧‧‧First translation memory

1021‧‧‧ original information

1021A‧‧‧ original clause

1022‧‧‧Translation materials

1022A‧‧‧ translation clause

103‧‧‧Comparative analysis unit

103R‧‧‧ analysis results

104‧‧‧Integrated unit

1041‧‧‧ Approximate translation memory

1042‧‧‧

105‧‧‧Substitute unit

1051‧‧‧The third translation memory

106‧‧‧Exporting unit

12‧‧‧Electronic devices

120‧‧‧Translated documents

120A‧‧‧Sentences to be translated

121‧‧‧Display interface

122‧‧‧storage unit

FIG. 1 is a schematic diagram of a translation memory enhancement system 1 of the present embodiment. FIG. 2 is a schematic diagram showing the analysis result 103R displayed on the display interface 121.

Claims (6)

A translation memory enhancement system, comprising: at least one electronic device, the electronic device comprising at least one file to be translated and a storage unit, the file to be translated is composed of a plurality of to-be-translated clauses; and a back-end server, The electronic device is connected to the backend server by means of a network connection, and the backend server comprises: a historical translation memory, the historical translation memory comprising a plurality of different first translation memories, the first The translation memory includes an original text and a translation data, the original data is composed of a plurality of original clauses, the translation data is composed of multiple translation clauses, and each of the original clauses corresponds to one of the translation clauses And a comparison analysis unit, the comparison analysis unit compares each of the to-be-translated clauses with all the original clauses of each original text to search for the original clause adjacent to the to-be-translated clause; Wherein, the comparison analysis unit approximates the original clause of the clause to be translated and the translation clause corresponding to the original clause Form an approximate translation memory. The translation memory enhancement system of claim 1, wherein the backend server further comprises an integration unit; when the approximate translation memory has sentences of the same or similar content, the integration unit has the same Or the sentences of similar content are integrated to transform the approximate translation memory into a second translation memory. The translation memory enhancement system of claim 2, wherein the backend server further comprises a replacement unit; when the second translation memory includes at least one brand text, the replacement unit performs the brand text Substitution to convert the second translation memory into a third translation memory. The translation memory enhancement system of claim 3, wherein the backend server further comprises a export unit, and the export unit exports the third translation memory to the storage unit. The translation memory enhancement system according to claim 1, wherein the electronic device further comprises a display interface, and the original text clauses are organized into an analysis result according to the degree of approximating the to-be-translated clause, and the analysis result is The comparison analysis unit transmits to the display interface. The translation memory enhancement system according to claim 1, wherein the electronic device is a desktop computer, a notebook computer, a tablet computer or a smart phone.