A method and related apparatus for machine translation of bantu place names

Abstract

The application discloses a Bantu language geographical name machine translation method and related device, and relates to the field of language translation. The method comprises the following steps: adopting a forward maximum matching algorithm and a reverse maximum matching algorithm to divide a Bantu language geographical name word group to be translated into letter combinations, and respectively calculating mutual information values of the obtained corresponding forward letter combination division result and reverse letter combination division result; determining an intermediate division result according to the comparison result of the mutual information values, and using a Bantu language Chinese translation and writing table to transliterate the intermediate division result to obtain a transliteration result. The application cuts the word group to be translated, obtains the intermediate division result which can be directly compared with the Bantu language Chinese translation and writing table for accurate proper name transliteration, and has high correlation between adjacent word groups, thereby solving the problem of proper name transliteration in Bantu language geographical name translation, enabling the machine translation model to independently perform Bantu language geographical name translation, reducing the human consumption during geographical name generation, and improving the machine translation efficiency.

Description

Technical Field

[0001] This application relates to the field of language translation, and in particular to a method and apparatus for machine translation of Bantu place names. Background Technology

[0002] Place name translation refers to converting the expression of a geographical entity in one language into its expression in another. Automatic place name translation, also known as machine place name translation, is a part of named entity translation within machine translation. However, this part is the most challenging of all named entity translations because: firstly, place names are complex, with significant differences in their composition across different geographical scales; secondly, translating Bantu place names requires considering proper noun transliteration, that is, transcribing proper nouns such as personal names, place names, and country names using Chinese characters with similar pronunciations. Currently, there are no dedicated automatic Bantu place name translation products in the industry, and general machine translation models cannot independently solve the problem of translating Bantu place names. Summary of the Invention

[0003] The purpose of this application is to provide a method and related apparatus for machine translation of Bantu place names, to solve the problem of proper noun transliteration in Bantu place name translation, so that the machine translation model can independently translate Bantu place names.

[0004] To achieve the above objectives, this application provides the following solution.

[0005] Firstly, this application provides a method for machine translation of Bantu place names, including the following:

[0006] The forward maximum matching algorithm and the backward maximum matching algorithm are used to divide the Bantu place name groups to be translated into letter combinations, so as to obtain the forward letter combination division results and the backward letter combination division results corresponding to the Bantu place name groups to be translated.

[0007] Calculate a first mutual information value and a second mutual information value, and compare the first mutual information value and the second mutual information value to obtain a first comparison result. When the first comparison result indicates that the first mutual information value is equal to the second mutual information value, select the forward letter combination segmentation result or the reverse letter combination segmentation result as an intermediate segmentation result. When the first comparison result indicates that the first mutual information value is greater than the second mutual information value, determine the forward letter combination segmentation result as an intermediate segmentation result. When the first comparison result indicates that the first mutual information value is less than the second mutual information value, determine the reverse letter combination segmentation result as an intermediate segmentation result. The first mutual information value is the mutual information value of the forward letter combination segmentation result; the second mutual information value is the mutual information value of the reverse letter combination segmentation result.

[0008] The intermediate segmentation results are transliterated using the Bantu-Chinese transliteration table to obtain the transliteration result; the Bantu-Chinese transliteration table includes multiple Bantu segmentation strings and the corresponding Chinese character transliteration strings for each Bantu segmentation string.

[0009] In a second aspect, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the Bantu place name machine translation method described in any one of the above.

[0010] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the Bantu place name machine translation method described above.

[0011] Fourthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the Bantu place name machine translation method described above.

[0012] According to the specific embodiments provided in this application, this application has the following technical effects.

[0013] This application provides a method and related apparatus for machine translation of Bantu place names. First, the Bantu place name group to be translated is obtained for subsequent processing. Second, the Bantu place name group to be translated is divided into letter combinations using a forward maximum matching algorithm and a backward maximum matching algorithm, resulting in forward letter combination division results and backward letter combination division results corresponding to the Bantu place name group to be translated. A first mutual information value and a second mutual information value are calculated, and the first mutual information value and the second mutual information value are compared to obtain a first comparison result. When the first comparison result indicates that the first mutual information value is equal to the second mutual information value, either the forward letter combination division result or the backward letter combination division result can be selected as an intermediate division result. When the first comparison result indicates that the first mutual information value is equal to the second mutual information value, the first comparison result indicates that the first mutual information value is equal to the second mutual information value. When the first mutual information value is greater than the second mutual information value, the forward letter combination segmentation result is determined as the intermediate segmentation result. When the first comparison result indicates that the first mutual information value is less than the second mutual information value, the reverse letter combination segmentation result is determined as the intermediate segmentation result. The first mutual information value is the mutual information value of the forward letter combination segmentation result. The second mutual information value is the mutual information value of the reverse letter combination segmentation result. Through the above steps, the Bantu place name group to be translated is divided into multiple word strings. These word strings have corresponding Chinese characters in the Bantu Chinese transliteration table, and the word strings have a high degree of correlation. Therefore, this step provides the conditions for accurate transliteration of proper names by referring to the Bantu Chinese transliteration table. Finally, the intermediate segmentation result is transliterated using the Bantu Chinese transliteration table to obtain the transliteration result. This invention segments the words to be translated, obtaining intermediate segmentation results that can be directly compared with the Bantu-Chinese transliteration table for accurate proper noun transliteration, and with high correlation between adjacent words. This solves the problem of proper noun transliteration in Bantu place name translation, enabling machine translation models to independently translate Bantu place names, reducing the human effort required for place name generation, and improving machine translation efficiency. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 This is a diagram illustrating the application environment of a Bantu language place name machine translation method according to an embodiment of this application.

[0016] Figure 2 This is a flowchart illustrating a Bantu language place name machine translation method provided in one embodiment of this application.

[0017] Figure 3This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0018] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0019] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0020] The Bantu place name machine translation method provided in this application embodiment can be applied to, for example... Figure 1In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be set up independently, integrated into server 104, or placed in the cloud or on other servers. Terminal 102 can send the Bantu place names to be translated to server 104. Server 104 receives the Bantu place names to be translated and, for each Bantu place name, uses a forward maximum matching algorithm and a reverse maximum matching algorithm to perform letter combination segmentation on the Bantu place names to be translated, obtaining the forward letter combination segmentation result and the reverse letter combination segmentation result corresponding to the Bantu place names to be translated; calculates a first mutual information value and a second mutual information value, and compares the first mutual information value and the second mutual information value to obtain a first comparison result. When the first comparison result indicates that the first mutual information value is equal to the second mutual information value, the forward letter combination segmentation result or the reverse letter combination segmentation result is selected. The parent combination segmentation result is used as an intermediate segmentation result. When the first comparison result indicates that the first mutual information value is greater than the second mutual information value, the forward letter combination segmentation result is determined as the intermediate segmentation result. When the first comparison result indicates that the first mutual information value is less than the second mutual information value, the reverse letter combination segmentation result is determined as the intermediate segmentation result. The first mutual information value is the mutual information value of the forward letter combination segmentation result. The second mutual information value is the mutual information value of the reverse letter combination segmentation result. The intermediate segmentation result is transliterated using the Bantu-Chinese transliteration table to obtain the transliteration result. The Bantu-Chinese transliteration table includes multiple Bantu segmentation strings and the corresponding Chinese transliteration strings for each Bantu segmentation string. The server 104 can feed back the obtained transliteration result to the terminal 102. In addition, in some embodiments, the Bantu place name machine translation method can also be implemented by the server 104 or the terminal 102 separately. For example, the terminal 102 can directly perform Bantu place name machine translation on the Bantu place name group to be translated, or the server 104 can obtain the Bantu place name group to be translated from the data storage system and perform Bantu place name machine translation on the Bantu place name group to be translated.

[0021] The terminal 102 can be, but is not limited to, various desktop computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, and smart in-vehicle devices. Portable wearable devices can include smartwatches, smart bracelets, and head-mounted devices. The server 104 can be implemented using a standalone server or a server cluster composed of multiple servers, or it can be a cloud server.

[0022] It should be noted that the reverse maximum matching algorithm is the reverse maximum matching algorithm during computer execution.

[0023] In one exemplary embodiment, such as Figure 2 As shown, a method for machine translation of Bantu place names is provided. This method is executed by a computer device, specifically by a terminal or server alone, or by both a terminal and a server. In this embodiment, the method is applied to... Figure 1 Taking server 104 as an example, the explanation includes the following steps 201 to 208.

[0024] Step 201: Obtain the Bantu place noun phrases to be translated.

[0025] Step 202: The forward maximum matching algorithm and the reverse maximum matching algorithm are used to divide the Bantu place name group to be translated into letter combinations, so as to obtain the forward letter combination division result and the reverse letter combination division result corresponding to the Bantu place name group to be translated.

[0026] Step 203: Calculate the first mutual information value and the second mutual information value, and compare the first mutual information value and the second mutual information value to obtain a first comparison result. When the first comparison result indicates that the first mutual information value is equal to the second mutual information value, select the forward letter combination segmentation result or the reverse letter combination segmentation result as the intermediate segmentation result. When the first comparison result indicates that the first mutual information value is greater than the second mutual information value, determine the forward letter combination segmentation result as the intermediate segmentation result. When the first comparison result indicates that the first mutual information value is less than the second mutual information value, determine the reverse letter combination segmentation result as the intermediate segmentation result. The first mutual information value is the mutual information value of the forward letter combination segmentation result; the second mutual information value is the mutual information value of the reverse letter combination segmentation result.

[0027] Step 204: Translate the intermediate segmentation results using the Bantu-Chinese transliteration table to obtain the transliteration results; the Bantu-Chinese transliteration table includes multiple Bantu segmentation strings and the corresponding Chinese character transliteration strings for each Bantu segmentation string.

[0028] By implementing steps 201 to 204 above, an intermediate segmentation result was obtained that can be directly compared with the Bantu Chinese transliteration table for accurate proper noun transliteration, and the correlation between adjacent word groups is high. This solves the problem of proper noun transliteration in Bantu place name translation, enabling the machine translation model to independently translate Bantu place names, reducing the human consumption in place name generation, and improving machine translation efficiency.

[0029] In another exemplary embodiment of this application, in order to optimize the transliteration result and obtain an optimized transliteration result that conforms to the national standard, the above step 204 is replaced by the following step 301.

[0030] Step 301: Optimize the transliteration result by using the Bantu place name instance library to obtain an optimized transliteration result; the Bantu place name instance library includes multiple Chinese character transliteration strings and the common names corresponding to each Chinese character transliteration string.

[0031] In the national standard, it is stipulated that for the Bantu place name phrase to be translated, if there is a fixed Chinese character translation, and this fixed Chinese character translation has been habitually used by large-scale publications in China for certain overseas place names for a long time, then the fixed Chinese character translation is adopted. For example, for "lesotho" which already has a fixed Chinese character translation "Lesotho", then directly adopt this fixed Chinese character translation "Lesotho" during translation.

[0032] In another exemplary embodiment of the present application, the Bantu place name phrase to be translated in step 201 is a phrase obtained after preprocessing the Bantu place name phrase input by the user.

[0033] In another exemplary embodiment of the present application, for the division of letter combinations, it is necessary to select letter combinations with as few included phrases as possible to meet the requirement of as short a transliteration as possible. To obtain the forward letter combination division result corresponding to the Bantu place name phrase to be translated, in the above step 202, the process of obtaining the forward letter combination division result corresponding to the Bantu place name phrase to be translated by using the forward maximum matching algorithm is replaced by the following steps 301 to 303.

[0034] Step 301: Determine the Bantu place name phrase to be translated as the first target phrase, and judge whether the length of the first target phrase is greater than 1 to obtain a first judgment result. When the first judgment result is negative, determine the first target phrase as the forward letter combination division result. When the first judgment result is positive, determine the first letter in the first target phrase as the first segmentation result, and perform the first operation.

[0035] Step 302, the first operation is as follows: determine whether the Bantu-Chinese translation table contains the first segmentation result, obtain a second judgment result, and when the second judgment result indicates yes, add a segmentation mark to the first segmentation result and update the remaining word groups in the first target word group without segmentation marks to the first target word group, and take the first letter in the first target word group to update the first segmentation result, and then perform the second operation on the first target word group. When the second judgment result indicates no, remove the first segmentation result in the first target word group, take the first letter in the remaining word group and concatenate it with the first segmentation result in the order of the Bantu place name group to be translated to generate a first concatenated word group, and update the first concatenated word group to the first segmentation result, and then update the word groups in the first target word group that do not contain the first segmentation result and do not contain segmentation marks to the first target word group, and finally perform the second operation on the first target word group.

[0036] Step 303, the second operation is: determine whether the length of the first target word group is greater than 1, obtain the third judgment result, and when the third judgment result is no, concatenate the word group containing the segmentation mark, the first segmentation result and the first target word group to generate the second concatenated word group, and determine the second concatenated word group as the positive letter combination division result, and when the third judgment result is yes, perform the first operation on the first segmentation result.

[0037] The operation of adding a segmentation marker in step 302 is to add a " / " character after the first segmentation result.

[0038] The Bantu-Chinese transliteration table used in the above steps was formulated with reference to the Bantu transliteration rules and is used as a dictionary.

[0039] In another exemplary embodiment of this application, in order to obtain the reverse letter combination segmentation result corresponding to the Bantu place name group to be translated, the process of obtaining the reverse letter combination segmentation result corresponding to the Bantu place name group to be translated using the reverse maximum matching algorithm in step 202 is replaced by the following steps 401 to 403.

[0040] Step 401: Identify the Bantu place name group to be translated as the second target word group, and determine whether the length of the second target word group is greater than 1 to obtain the fourth judgment result. When the fourth judgment result is negative, identify the second target word group as the reverse letter combination segmentation result. When the fourth judgment result is positive, identify the last letter in the second target word group as the second segmentation result and perform the third operation.

[0041] Step 402, the third operation is as follows: determine whether the Bantu-Chinese translation table contains the second segmentation result, obtain the fifth judgment result, and when the fifth judgment result indicates yes, add the second segmentation mark to the second segmentation result, update the remaining word groups in the second target word group without segmentation marks to the second target word group, and take the last letter in the second target word group to update the second segmentation result, and then perform the fourth operation on the second target word group. When the fifth judgment result indicates no, remove the second segmentation result in the second target word group, take the last letter in the remaining word group and concatenate it with the second segmentation result in the order of the Bantu place name group to be translated to generate the third concatenated word group, and update the third concatenated word group to the second segmentation result, then update the word groups in the second target word group that do not contain the second segmentation result and do not contain segmentation marks to the second target word group, and finally perform the fourth operation on the second target word group.

[0042] Step 403, the fourth operation is: determine whether the length of the second target word group is greater than 1, obtain the sixth judgment result, and when the sixth judgment result is no, concatenate the word group containing the segmentation mark, the second segmentation result and the second target word group to generate the fourth concatenated word group, and determine the fourth concatenated word group as the reverse letter combination division result; when the sixth judgment result is yes, perform the third operation on the second segmentation result.

[0043] The operation of adding a segmentation marker in step 402 is to add a " / " character before the second segmentation result.

[0044] In another exemplary embodiment of this application, in order to obtain the reverse letter combination segmentation result corresponding to the Bantu place name group to be translated, the step of calculating the first mutual information value and the second mutual information value in step 203 above is replaced by the following steps 501 to 502.

[0045] Step 502: Calculate the average value of the mutual information between any two adjacent word groups in the positive letter combination segmentation result, and use it as the first mutual information value.

[0046] Step 502: Calculate the average value of the mutual information between any two adjacent word groups in the reverse letter combination segmentation result, and use it as the second mutual information value.

[0047] In the above steps, mutual information represents the degree of correlation between two statistical data. The higher the correlation, the more mutual information; the lower the correlation, the less mutual information. In this embodiment, the first mutual information value refers to the average value of the mutual information between any two adjacent letter strings in the forward letter combination segmentation result and the average value of the mutual information between all adjacent letter strings. The second mutual information value refers to the average value of the mutual information between any two adjacent letter strings in the reverse letter combination segmentation result and the average value of the mutual information between all adjacent letter strings. The second mutual information value refers to the average value of the mutual information between any two adjacent letter strings in the reverse letter combination segmentation result and the average value of the mutual information between all adjacent letter strings.

[0048] In another exemplary embodiment of this application, in order to ensure the correctness of the format of the obtained Bantu place name group to be translated, the Bantu place name group input by the user can be preprocessed. Before step 201 above, the method may also include the following steps 601 to 605.

[0049] Step 601: Determine whether the Bantu place name group entered by the user contains illegal characters to obtain the seventh judgment result. If the seventh judgment result is yes, delete all illegal characters in the Bantu place name group entered by the user to obtain the first preprocessing result. If the seventh judgment result is no, use the Bantu place name group entered by the user as the first preprocessing result.

[0050] Step 602: Determine whether the first preprocessing result contains numbers or strings containing numbers to obtain the eighth determination result; if the eighth determination result is yes, replace the numbers in the first preprocessing result with the corresponding Chinese characters to obtain the second preprocessing result; if the eighth determination result is no, use the first preprocessing result as the second preprocessing result.

[0051] Step 603: Detect whether the second preprocessing result contains a date string to obtain the ninth judgment result. If the ninth judgment result is yes, replace the date string with the corresponding legal date string to obtain the third preprocessing result; if the ninth judgment result is no, use the second preprocessing result as the third preprocessing result; the writing order of the legal date string is month, day, year.

[0052] Step 604: Detect whether the third preprocessing result contains an article to obtain the tenth judgment result. If the tenth judgment result is yes, delete the article to obtain the Bantu place name group to be translated; if the tenth judgment result is no, use the third preprocessing result as the fourth preprocessing result.

[0053] Step 605: Determine if the letter 'm' exists in the fourth preprocessing result to obtain the eleventh judgment result; if not, use the fourth preprocessing result as the Bantu place name group to be translated; if yes, determine if 'n' or 'p' exists in the fourth preprocessing result to obtain the twelfth judgment result; if the twelfth judgment result is no, use the fourth preprocessing result as the Bantu place name group to be translated; if the twelfth judgment result is yes, determine if the letter 'm' is adjacent to 'b' or 'p' and precedes 'b' or 'p' to obtain the thirteenth judgment result; if the thirteenth judgment result is yes, replace the letter 'm' in the fourth preprocessing result with 'n' to obtain the Bantu place name group to be translated; if the thirteenth judgment result is no, use the fourth preprocessing result as the Bantu place name group to be translated.

[0054] In step 604, the article is an article string in the Bantu writing system.

[0055] Regarding steps 601 to 605 above, this application provides an example as follows:

[0056] Because the Bantu place names entered by users may contain errors in place name format or articles, which are detrimental to machine translation, it is necessary to preprocess the Bantu place name groups entered by users in advance for subsequent translation.

[0057] The preprocessing includes the following steps:

[0058] 1. Determine if the user-input Bantu place name group contains illegal characters and delete them. Illegal characters such as “” and “&” are used to obtain the first preprocessing result.

[0059] 2. Determine whether the first preprocessing result contains numbers or strings containing numbers. Modify the writing style of the numbers and replace them with the corresponding Chinese characters. For example, write the forms 10-1, 10 / 2, etc. in accordance with the writing style of Chinese place names to obtain the second preprocessing result.

[0060] 3. Determine whether the second preprocessing result contains a date string. The date format should be uniformly written in the format of month, day, and year to obtain the third preprocessing result.

[0061] 4. Determine whether the third preprocessing result contains articles. If so, delete the articles and obtain the fourth preprocessing result.

[0062] 5. Determine if the letter 'm' exists in the fourth preprocessing result. If it does, determine if 'm' precedes the letters 'b' or 'p'. If so, rewrite 'm' as 'n' in the fourth preprocessing result to obtain the Bantu place name group to be translated. This step optimizes the transliteration result based on transliteration rules. The national standard provides relevant regulations for some letters and their transliterations. Transliteration optimization based on transliteration rules mainly identifies specific rules for letter combinations in word letters, and then obtains the letter combination translation result from the letter combination rule tuple, thereby obtaining the transliterated word's Chinese character transliteration result. For example, the standard stipulates that the consonant letter 'm' before the letters 'b' or 'p' must be transliterated as [n]. Therefore, when transliterating words involving this rule, 'm' needs to be changed to 'n' before transliteration.

[0063] To more clearly illustrate the specific solution of the present invention, an example is given below for detailed explanation.

[0064] Get the Bantu place noun phrase "Westdene" to be translated.

[0065] The Bantu place name phrase “Westdene” to be translated is divided into positive letter combinations using a forward maximum matching algorithm: the Bantu place name phrase to be translated is identified as the first target phrase, and it is determined whether the length of the first target phrase is greater than 1. If its letter string length is detected to be 8, that is, the length is greater than 1, then the first letter “w” is extracted from “Westdene”, and “w” is determined as the first segmentation result. The first operation is then performed, that is, it is determined whether the Bantu-Chinese transliteration table contains the first segmentation result “w”. If the first segmentation result "w" is not found in the Bantu-Chinese transliteration table, then after removing the first segmentation result "w" from the first target phrase "Westdene", the first letter of the remaining phrase is concatenated with the first segmentation result in the order of the Bantu place names to be translated, generating the first concatenated phrase "we". This concatenated phrase is then updated as the first segmentation result. Next, the phrase "stdene", which does not contain the first segmentation result and has no segmentation marker, is updated as the first target phrase. The second operation is then performed on this phrase, namely, determining whether the length of the first target phrase "stdene" is greater than 1. Since the length of the first target phrase "stdene" is determined to be 6, which is greater than 1, the first operation is performed on the first segmentation result "we", namely, determining whether the first segmentation result is present in the Bantu-Chinese transliteration table. If the Bantu-Chinese transliteration table contains "we", then add a segmentation marker "we / " to the first segmentation result "we". Update the remaining unmarked word groups in the first target word group "stdene" to the first target word group "stdene". Perform the second operation. If the length of "stdene" is greater than 1, take the first letter "s" to update the first segmentation result. Perform the first operation. If the Bantu-Chinese transliteration table contains "s", then mark "s" as "s / ". Perform the second operation again. Segment the marked word groups. Take the remaining word group "tdene" as the first target word group. Perform the second operation. Determine the length of the remaining letter string. If the length is greater than 1, perform the first operation. Continue to segment the first letter "t" from the first target word group "tdene" and continue processing in the above way until the length of the remaining letter string is 1. The final positive letter combination segmentation result is "we / s / t / den / e".

[0066] For the Bantu place names to be translated, a reverse maximum matching algorithm is used for reverse letter combination segmentation: "Westdene" is obtained and identified as the second target phrase. Its length is checked and found to be 8, which is greater than 1. If so, the last letter "e" in the third target phrase is extracted as the second segmentation result. The third operation is then performed: checking if the Bantu-Chinese transliteration table contains "e". If the result is negative, the extracted second segmentation result "e" from "Westdene" is removed, "Westden" is updated as the second target phrase, and its last letter is concatenated with the second segmentation result in sequence to generate the third concatenated phrase "ne". "ne" is then updated as the second segmentation result, and the second target phrase is updated to "Westde". The fourth operation is then performed: checking its length. If its length is 6, which is greater than 1, the third operation is performed: checking if the Bantu-Chinese transliteration table contains "ne". If the judgment result is yes, then add a segmentation marker, " / ne", to the second segmentation result, extract the last letter "e" from "Westde" to update the second segmentation result, update the extracted "Westd" as the second target phrase, and perform the fourth operation, which is to judge the length of the second target phrase. If the length is 5, which is greater than 1, then perform the third operation, which is to judge whether the Bantu Chinese transliteration table contains "e". If the judgment result is no, then segment the last letter "d" from the second target phrase "Westd", combine it with the second segmentation result "e" to form the third concatenation phrase "de", update "de" to the second segmentation result, update the second target phrase to "West", and perform the fourth operation, which is to judge whether the length of the second target phrase is greater than 1. If the length is greater than 1, then continue to process it in the above way until the length of the remaining letter string is 1. The final reverse letter combination segmentation result is "we / s / t / de / ne".

[0067] Calculate the mutual information values ​​between "we" and "s", "s" and "t", "t" and "den", and "den" and "e" in the forward letter combination segmentation result "we / s / t / den / e". Then calculate the average of these five mutual information values ​​and record this average as the first mutual information value. Next, calculate the mutual information values ​​between "we" and "s", "s" and "t", "t" and "de", and "de" and "ne" in the reverse letter combination segmentation result "we / s / t / de / ne". Finally, calculate the average of these five mutual information values ​​and record this average as the second mutual information value.

[0068] Comparing the first mutual information value and the second mutual information value, it is found that the first mutual information value is less than the second mutual information value. Therefore, the reverse letter combination partition result "we / s / t / de / ne" is selected as the intermediate partition result of the place name group "westdene".

[0069] The intermediate division results were transliterated using the Bantu-Chinese transliteration table to obtain the transliteration results.

[0070] Based on the same inventive concept, this application also provides a Bantu place name machine translation device for implementing the Bantu place name machine translation method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations of one or more Bantu place name machine translation device embodiments provided below can be found in the limitations of the Bantu place name machine translation method described above, and will not be repeated here.

[0071] Those skilled in the art will understand that Figure 3 The structures shown are merely block diagrams of some structures related to the present application and do not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than shown in the figures, or combine certain components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0072] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0073] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0074] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Moreover, the collection, use and processing of the relevant data are carried out in compliance with the relevant data protection laws and policies of the country where the location is located, and with the authorization granted by the owner of the corresponding device.

[0075] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0076] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0077] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0078] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for machine translation of Bantu place names, characterized in that, The Bantu place name machine translation method includes: Obtain the Bantu place names to be translated; The forward maximum matching algorithm and the backward maximum matching algorithm are used to divide the Bantu place name group to be translated into letter combinations, and the forward letter combination division results and the backward letter combination division results corresponding to the Bantu place name group to be translated are obtained. Calculate a first mutual information value and a second mutual information value, and compare the first mutual information value and the second mutual information value to obtain a first comparison result. When the first comparison result indicates that the first mutual information value is equal to the second mutual information value, select the forward letter combination segmentation result or the reverse letter combination segmentation result as an intermediate segmentation result. When the first comparison result indicates that the first mutual information value is greater than the second mutual information value, determine the forward letter combination segmentation result as an intermediate segmentation result. When the first comparison result indicates that the first mutual information value is less than the second mutual information value, determine the reverse letter combination segmentation result as an intermediate segmentation result. The first mutual information value is the mutual information value of the forward letter combination segmentation result; the second mutual information value is the mutual information value of the reverse letter combination segmentation result. The intermediate segmentation results are transliterated using the Bantu-Chinese transliteration table to obtain the transliteration result; the Bantu-Chinese transliteration table includes multiple Bantu segmentation strings and the corresponding Chinese character transliteration strings for each Bantu segmentation string; The Bantu place name group to be translated is a group of words obtained after preprocessing the Bantu place name group input by the user. The preprocessing process is as follows: Determine whether the user-input Bantu place name group contains illegal characters to obtain the seventh judgment result. If the seventh judgment result is yes, delete all illegal characters in the user-input Bantu place name group to obtain the first preprocessing result; if the seventh judgment result is no, use the user-input Bantu place name group as the first preprocessing result. Determine whether the first preprocessing result contains numbers or strings containing numbers to obtain the eighth determination result; if the eighth determination result is yes, replace the numbers in the first preprocessing result with the corresponding Chinese characters to obtain the second preprocessing result; if the eighth determination result is no, use the first preprocessing result as the second preprocessing result. The system determines whether the second preprocessing result contains a date string, resulting in a ninth determination. If the ninth determination is yes, the date string is replaced with the corresponding valid date string, resulting in a third preprocessing result. If the ninth determination is no, the second preprocessing result is used as the third preprocessing result. The valid date string is written in the order of month, day, and year. Determine whether the third preprocessing result contains an article to obtain the tenth judgment result. If the tenth judgment result is yes, delete the article to obtain the fourth preprocessing result; if the tenth judgment result is no, use the third preprocessing result as the fourth preprocessing result. The system determines whether the letter 'm' exists in the fourth preprocessing result, resulting in the eleventh judgment. If not, the fourth preprocessing result is used as the Bantu place name group to be translated. If yes, the system determines whether 'n' or 'p' exists in the fourth preprocessing result, resulting in the twelfth judgment. If the twelfth judgment is no, the fourth preprocessing result is used as the Bantu place name group to be translated. If the twelfth judgment is yes, the system determines whether the letter 'm' is adjacent to 'b' or 'p' and precedes 'b' or 'p', resulting in the thirteenth judgment. If the thirteenth judgment is yes, the letter 'm' in the fourth preprocessing result is replaced with 'n', resulting in the Bantu place name group to be translated. If the thirteenth judgment is no, the fourth preprocessing result is used as the Bantu place name group to be translated.

2. The Bantu place name machine translation method according to claim 1, characterized in that, The transliteration result is a Chinese character transliteration string, which also includes: The transliteration results are optimized using a Bantu place name instance library to obtain optimized transliteration results; the Bantu place name instance library includes various Chinese character transliteration strings and the general translation name corresponding to each Chinese character transliteration transliteration string.

3. The Bantu place name machine translation method according to claim 1, characterized in that, The process for determining the positive letter combination segmentation results corresponding to the Bantu place name groups to be translated is as follows: The first target word group is identified as the Bantu place name group to be translated, and it is determined whether the length of the first target word group is greater than 1. The first judgment result is obtained. When the first judgment result is negative, the first target word group is identified as the positive letter combination segmentation result. When the first judgment result is positive, the first letter in the first target word group is identified as the first segmentation result, and the first operation is performed. The first operation is as follows: determine whether the Bantu-Chinese transliteration table contains the first segmentation result, obtain a second judgment result, and when the second judgment result indicates yes, add a segmentation mark to the first segmentation result and update the remaining word groups in the first target word group without segmentation marks to the first target word group, and take the first letter in the first target word group to update the first segmentation result, and then perform the second operation on the first target word group. When the second judgment result indicates no, remove the first segmentation result in the first target word group, take the first letter in the remaining word group and concatenate it with the first segmentation result in the order of the Bantu place name group to be translated to generate the first concatenated word group, and update the first concatenated word group to the first segmentation result, then update the word groups in the first target word group that do not contain the first segmentation result and do not contain segmentation marks to the first target word group, and finally perform the second operation on the first target word group. The second operation is as follows: determine whether the length of the first target word group is greater than 1, obtain the third judgment result, and when the third judgment result is negative, concatenate the word group containing the segmentation mark, the first segmentation result and the first target word group to generate the second concatenated word group, and determine the second concatenated word group as the positive letter combination segmentation result. When the third judgment result is positive, perform the first operation on the first segmentation result.

4. The Bantu place name machine translation method according to claim 1, characterized in that, The process for determining the reverse letter combination results corresponding to the Bantu place name groups to be translated is as follows: The Bantu place name group to be translated is identified as the second target word group, and it is determined whether the length of the second target word group is greater than 1 to obtain the fourth judgment result. When the fourth judgment result is negative, the second target word group is identified as the reverse letter combination segmentation result. When the fourth judgment result is positive, the last letter in the second target word group is identified as the second segmentation result, and the third operation is performed. The third operation is as follows: determine whether the Bantu-Chinese transliteration table contains the second segmentation result, obtain the fifth judgment result, and when the fifth judgment result indicates yes, add a segmentation mark to the second segmentation result, update the remaining word groups in the second target word group without segmentation marks to the second target word group, and take the last letter in the second target word group to update the second segmentation result, and then perform the fourth operation on the second target word group. When the fifth judgment result indicates no, remove the second segmentation result in the second target word group, take the last letter in the remaining word group and concatenate it with the second segmentation result in the order of the Bantu place name group to be translated to generate the third concatenated word group, and update the third concatenated word group to the second segmentation result, then update the word groups in the second target word group that do not contain the second segmentation result and do not contain segmentation marks to the second target word group, and finally perform the fourth operation on the second target word group. The fourth operation is as follows: determine whether the length of the second target word group is greater than 1, obtain the sixth judgment result, and when the sixth judgment result is negative, concatenate the word group containing the segmentation mark, the second segmentation result and the second target word group to generate the fourth concatenated word group, and determine the fourth concatenated word group as the reverse letter combination division result. When the sixth judgment result is positive, perform the third operation on the second segmentation result.

5. The Bantu place name machine translation method according to claim 1, characterized in that, The calculation of the first mutual information value and the second mutual information value specifically includes: Calculate the average mutual information value between any two adjacent word groups in the positive letter combination segmentation result, and use it as the first mutual information value; Calculate the average mutual information value between any two adjacent word groups in the reverse letter combination segmentation result, and use it as the second mutual information value.

6. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that the processor executes the computer program to implement the Bantu place name machine translation method according to any one of claims 1-5.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the Bantu place name machine translation method as described in any one of claims 1-5.

8. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the Bantu place name machine translation method as described in any one of claims 1-5.

Images