A method for automatic translation of place names based on a specific rule base

Abstract

The present application relates to the field of geographical name translation technology, and specifically discloses a geographical name automatic translation method based on a specific rule base, which first splits the geographical name to be translated input by the user to obtain multiple geographical name segments, and loads the corresponding rule base based on the type of each geographical name segment to parse each geographical name segment to obtain multiple initial candidate translated names, and then introduces a natural language processing technology based on deep learning to perform semantic analysis and contextual semantic association encoding on each initial candidate translated name to obtain the global contextual semantic association features of the candidate translated name, and with the help of a large language model, the geographical name is translated based on the contextual semantic information of the candidate translated name to generate a geographical name translation result. By combining a specific rule base and semantic analysis technology, this method can comprehensively consider the language characteristics, cultural background and contextual context of the geographical name, thereby improving the accuracy of geographical name translation.

Description

Technical Field

[0001] The present application relates to the technical field of place name translation, and more specifically, to a method for automatic place name translation based on a specific rule base. Background Art

[0002] With the acceleration of globalization, cross-border exchanges are becoming more frequent. Place names, as an important part of geographic information, play an indispensable role in international exchanges, tourism, logistics and other fields. Place name translation is an important link in cross-cultural communication and geographic information processing. It requires accurate and appropriate conversion of place names in one language into another language while maintaining the meaning, cultural characteristics and geographical orientation of the place names.

[0003] However, traditional place name translation methods often rely on manual translation or automatic translation based on simple rules. On the one hand, manual translation is inefficient. Faced with the massive demand for place name translation, manual translation often cannot meet the timeliness requirements, seriously affecting the project progress. On the other hand, although the rule-based translation method has improved the translation efficiency to a certain extent, the formulation of rules is often based on limited language samples. In reality, place names come from a wide range of sources and have complex and diverse language structures. Many place names have unique historical, cultural, and geographical backgrounds, which are difficult to fully cover with rules, resulting in insufficient accuracy and cultural adaptability of rule-based translation results.

[0004] Therefore, an optimized method for automatic translation of place names based on a specific rule base is needed to solve the above technical problems. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a method for automatic translation of place names based on a specific rule base, which first splits the place name to be translated input by the user to obtain multiple place name fragments, and loads the corresponding rule base based on the type of each place name fragment to parse each place name fragment to obtain multiple initial candidate translated names, and then introduces a natural language processing technology based on deep learning to perform semantic analysis and contextual semantic association encoding on each initial candidate translated name to obtain the global contextual semantic association features of the candidate translated name, and with the help of a large language model, the place name is translated based on the contextual semantic information of the candidate translated name to generate a place name translation result. By combining a specific rule base and semantic parsing technology, this method can comprehensively consider the language characteristics, cultural background and contextual context of the place name, thereby improving the accuracy of place name translation.

[0006] According to one aspect of the present application, a method for automatically translating place names based on a specific rule base is provided, which includes: Obtaining the place name to be translated input by the user; Splitting the place name to be translated to obtain a sequence distribution of place name fragments; Based on the type of each place name segment in the sequence distribution of the place name segment, the each place name segment is parsed based on a specific rule base to obtain a sequence distribution of initial candidate translated names; Performing semantic embedding coding on each initial candidate translated name in the sequence distribution of the initial candidate translated names to obtain a sequence distribution of semantic embedding coding vectors of the initial candidate translated names; Performing context-association coding based on semantic change perception on the sequence distribution of the initial candidate translated name semantic embedding coding vector to obtain a candidate translated name context semantic association coding vector; Based on the candidate translated name context semantic association encoding vector, a place name translation result is generated.

[0007] Specifically, based on the type of each place name segment in the sequence distribution of the place name segment, the each place name segment is parsed based on a specific rule base to obtain the sequence distribution of the initial candidate translated name, including: Identifying the type of each place name segment in the sequence distribution of the place name segments based on the place name classifier to obtain a sequence distribution of the place name segment types; Based on the sequence distribution of the place name segment type, loading a rule library of the corresponding place name type; Based on the rule base of the corresponding place name type, each place name segment in the sequence distribution of the place name segments is parsed to obtain the sequence distribution of the initial candidate translated names.

[0008] Specifically, performing semantic embedding coding on each initial candidate translated name in the sequence distribution of the initial candidate translated names to obtain a sequence distribution of semantic embedding coding vectors of the initial candidate translated names includes: A semantic embedding encoder based on the Bert model is used to perform semantic embedding encoding on each initial candidate translated name in the sequence distribution of the initial candidate translated names to obtain a sequence distribution of the semantic embedding encoding vectors of the initial candidate translated names.

[0009] Specifically, performing context-related coding based on semantic change perception on the sequence distribution of the initial candidate translated name semantic embedding coding vector to obtain the candidate translated name context semantic association coding vector includes: Performing a context-neighborhood-aware semantic transfer significance measurement on each initial candidate translated name semantic embedding coding vector in the sequence distribution of the initial candidate translated name semantic embedding coding vector to obtain a sequence distribution of candidate translated name semantic transfer significance factors; Based on the sequence distribution of the semantic transfer significance factors of the candidate translated names, feature modulation transfer aggregation coding is performed on the sequence distribution of the initial candidate translated name semantic embedding coding vector to obtain the candidate translated name context semantic association coding vector.

[0010] Specifically, each initial candidate translated name semantic embedding coding vector in the sequence distribution of the initial candidate translated name semantic embedding coding vector is subjected to a contextual neighborhood-aware semantic transfer significance measurement to obtain a sequence distribution of candidate translated name semantic transfer significance factors, including: Calculating the semantic jump degree of each initial candidate translated name semantic embedding coding vector in the sequence distribution of the initial candidate translated name semantic embedding coding vector to obtain the sequence distribution of the candidate translated name semantic jump degree; Calculating the semantic transfer space span of each initial candidate translated name semantic embedding coding vector in the sequence distribution of the initial candidate translated name semantic embedding coding vector to obtain the sequence distribution of the candidate translated name semantic transfer space span; Based on the semantic jump degree and semantic transfer space span of each initial candidate translated name semantic embedding coding vector in the sequence distribution of the initial candidate translated name semantic embedding coding vector, the semantic transfer significance factor of each initial candidate translated name semantic embedding coding vector is calculated to obtain the sequence distribution of the candidate translated name semantic transfer significance factor.

[0011] Specifically, based on the sequence distribution of the semantic transfer significance factor of the candidate translated name, feature modulation transfer aggregation coding is performed on the sequence distribution of the initial candidate translated name semantic embedding coding vector to obtain the candidate translated name context semantic association coding vector, including: Inputting the sequence distribution of the semantic transfer significance factors of the candidate translated names into a gated transfer unit including a softmax normalization function to obtain the sequence distribution of the semantic transfer significance weights of the candidate translated names; Based on the sequence distribution of the candidate translated name semantically significant weights, weighted aggregation is performed on the sequence distribution of the initial candidate translated name semantic embedding coding vectors to obtain the candidate translated name context semantic association coding vector.

[0012] Specifically, based on the candidate translated name context semantic association encoding vector, a place name translation result is generated, including: After adding a prompt word to the tail of the candidate translated name context semantic association encoding vector, it is input into a translation polisher based on a large language model to obtain the place name translation result.

[0013] Specifically, the prompt word is: "polish the place name translation based on the contextual semantic information of the candidate translated names."

[0014] This application has at least the following technical effects: Compared with the prior art, the method for automatically translating place names based on a specific rule base provided by the present application first splits the place names to be translated input by the user to obtain multiple place name segments, and loads the corresponding rule base based on the type of each place name segment to parse each place name segment to obtain multiple initial candidate translated names, and then introduces natural language processing technology based on deep learning to perform semantic analysis and contextual semantic association encoding on each initial candidate translated name to obtain the global contextual semantic association features of the candidate translated names, and with the help of a large language model, the place name is translated based on the contextual semantic information of the candidate translated names to generate a place name translation result. By combining a specific rule base and semantic analysis technology, this method can comprehensively consider the language characteristics, cultural background and contextual context of place names, thereby improving the accuracy of place name translation. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0016] Figure 1 The present invention is a flowchart of a method for automatically translating place names based on a specific rule base according to an embodiment of the present application.

[0017] Figure 2 This is a data flow diagram of a method for automatic translation of place names based on a specific rule base according to an embodiment of the present application.

[0018] Figure 3 This is a flowchart of sub-step S3 of the method for automatic translation of place names based on a specific rule base according to an embodiment of the present application.

[0019] Figure 4 This is a flowchart of sub-step S5 of the method for automatic translation of place names based on a specific rule base according to an embodiment of the present application.

[0020] Figure 5 This is a flowchart of sub-step S51 of the method for automatic translation of place names based on a specific rule base according to an embodiment of the present application.

[0021] Figure 6 This is a flowchart of sub-step S52 of the method for automatic translation of place names based on a specific rule base according to an embodiment of the present application. DETAILED DESCRIPTION

[0022] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0023] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0024] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. Meanwhile, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.

[0025] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0026] It should be noted that all data acquisition actions in this application are carried out in compliance with the relevant data protection laws and policies of the country where the application is located, and with the authorization given by the owner of the corresponding device.

[0027] In response to the technical problems described in the above background technology, this application proposes an optimized method for automatic translation of place names based on a specific rule base, which first splits the place name to be translated input by the user to obtain multiple place name fragments, and loads the corresponding rule base based on the type of each place name fragment to parse each place name fragment to obtain multiple initial candidate translated names, and then introduces natural language processing technology based on deep learning to perform semantic analysis and contextual semantic association encoding on each initial candidate translated name to obtain the global contextual semantic association features of the candidate translated name, and with the help of a large language model, the place name is translated based on the contextual semantic information of the candidate translated name to generate a place name translation result. By combining a specific rule base and semantic parsing technology, this method can comprehensively consider the language characteristics, cultural background and contextual context of the place name, thereby improving the accuracy of place name translation.

[0028] Figure 1 The present invention is a flowchart of a method for automatically translating place names based on a specific rule base according to an embodiment of the present application. Figure 2Schematic diagram of data flow of the method for automatic translation of place names based on a specific rule base according to an embodiment of the present application. Figure 1 and Figure 2 As shown, the method for automatic translation of place names based on a specific rule base includes the following steps: S1, obtaining a place name to be translated input by a user; S2, splitting the place name to be translated to obtain a sequence distribution of place name segments; S3, based on the type of each place name segment in the sequence distribution of the place name segments, parsing each place name segment based on a specific rule base to obtain a sequence distribution of initial candidate translated names; S4, performing semantic embedding coding on each initial candidate translated name in the sequence distribution of the initial candidate translated names to obtain a sequence distribution of semantic embedding coding vectors of initial candidate translated names; S5, performing context association coding based on semantic change perception on the sequence distribution of the semantic embedding coding vectors of the initial candidate translated names to obtain a context semantic association coding vector of a candidate translated name; S6, generating a place name translation result based on the context semantic association coding vector of the candidate translated name.

[0029] In the above-mentioned method for automatically translating place names based on a specific rule base, the step S1 is to obtain the place name to be translated input by the user. It should be understood that the purpose of automatically translating place names is to meet the user's needs for translating specific place names. In practical applications, a special input box is usually set in the software interface or application, and the user can directly manually enter the place name to be translated in the input box. At the same time, in order to improve the convenience and accuracy of input, the system may provide some auxiliary functions, such as automatic association function. When the user enters partial characters, the system provides the user with possible complete place name options based on historical data or common place name databases, which is convenient for the user to quickly select; it may also be set to provide an error correction prompt function. When the content entered by the user has obvious spelling errors or does not conform to the conventional place name format, the system will promptly give prompts to guide the user to correct it.

[0030] Specifically, in order to adapt to the application needs in different scenarios, the system supports multiple ways to receive the place name information to be translated provided by users. On the one hand, for users who access the system through web pages or mobile applications, the interface design is intuitive and friendly, providing a concise and clear text input box, allowing users to directly type the place name to be translated. At the same time, in order to improve the user's input efficiency and reduce possible spelling errors, the system integrates an intelligent prompt function, which can quickly associate and display a list of possible target place names for selection based on some of the input characters. This not only speeds up the information entry speed, but also increases the accuracy of the initial input. In addition, the intelligent prompt function also combines historical search records and geographic location information to make the recommendation results more personalized and relevant. On the other hand, considering that some users may not know the correct way to write a specific place name, or want to find a specific place from a larger geographical range, the system also provides a map interaction function. Users can mark the target location on the map, and the system will automatically retrieve and provide the main place names in the nearby area for users to confirm based on the selected location. This method is particularly suitable for people who have special requirements for geographical location, such as travel enthusiasts and logistics practitioners. Even users who are not familiar with the local language can accurately specify the place name to be translated. The map interactive interface also provides rich geographic information query functions, such as surrounding facilities, transportation routes, etc., further enhancing the user experience.

[0031] In response to the needs of batch processing, such as the update of address databases of multinational companies or large-scale place name conversion tasks in academic research, the system has opened a file upload interface that supports the import of data files in various common formats such as Excel and CSV. Users can submit documents containing a large number of place name records at one time, and the system will parse and store these place names one by one, ready to enter the next step of the translation process. To ensure data security and privacy protection, an encrypted transmission protocol is used during the upload process, and relevant laws and regulations on the handling of personal information are strictly followed. In addition, in order to facilitate user management and tracking of the status of upload tasks, the system also provides detailed log records and progress reports.

[0032] The application of speech recognition technology has greatly broadened the service scope of the system, especially for users who are not good at typing or cannot use traditional input devices. The system is compatible with speech recognition in multiple languages ​​and dialects, and has strong noise resistance and speech speed adaptability, ensuring that the user's intention can be accurately captured even in noisy environments or when speaking at different speeds. The voice assistant can not only process query requests for a single place name, but also understand more complex natural language instructions, such as "find all the cities that the shortest path from Beijing to Paris passes through", and convert these instructions into a specific list of place names. Users can select from the list of place names to determine the place names to be translated.

[0033] In the above-mentioned method for automatic translation of place names based on a specific rule base, the step S2 is to split the place names to be translated to obtain a sequence distribution of place name fragments. It should be understood that, considering that place names are often composed of multiple parts with different semantic and grammatical functions, these parts may follow different translation rules. For example, a complete place name may contain different parts such as administrative division names, natural geographical entity names, landmark buildings or facility names, and different translation strategies may need to be adopted during translation. Therefore, the present application forms a sequence distribution of place name fragments by splitting the place names to be translated to decompose them into multiple place name fragments. In this way, each place name fragment corresponds to a component in the original place name, so that the corresponding rule base can be selected for each place name fragment for parsing, thereby improving the accuracy and flexibility of the translation.

[0034] Specifically, considering that place names may consist of multiple parts, such as city names, street names, building names or numbers, the system adopts a multi-level analysis method. For complex place names, the system first identifies the components within the overall framework, such as distinguishing different elements by detecting specific conjunctions or punctuation marks. This preliminary classification helps to break down complex place names into smaller, more manageable units. On this basis, the system applies a more detailed language model for further analysis for each individual component. For example, for place names that contain directional indications (east, west, south, north) or descriptive words (big, small, new, old), the system will identify these modifiers according to predefined patterns and separate the corresponding parts as independent place name fragments. In addition, the system can also identify and separate place name fragments with time attributes, such as "old" and "new", as well as adjectives that indicate size, such as "big" and "small".

[0035] In the above-mentioned method for automatic translation of place names based on a specific rule base, the step S3 is to parse each place name segment based on the type of each place name segment in the sequence distribution of the place name segment based on the specific rule base to obtain the sequence distribution of the initial candidate translation name. Figure 3 FIG. 1 is a flowchart of sub-step S3 of the method for automatically translating place names based on a specific rule base according to an embodiment of the present application. Figure 3 As shown, the step S3 includes the steps of: S31, identifying the type of each place name segment in the sequence distribution of the place name segment based on the place name classifier to obtain the sequence distribution of the place name segment type; S32, loading the rule library of the corresponding place name type based on the sequence distribution of the place name segment type; S33, parsing each place name segment in the sequence distribution of the place name segment based on the rule library of the corresponding place name type to obtain the sequence distribution of the initial candidate translated name.

[0036] Specifically, the step S31 identifies the type of each place name segment in the sequence distribution of the place name segment based on the place name classifier to obtain the sequence distribution of the place name segment type. It should be understood that different types of place name segments have different translation habits and rules. For example, the translation of administrative division names often follows the official translation standards, while the translation of natural geographical entity names may rely more on cultural habits and geographical features. Therefore, the present application further utilizes a place name classifier to identify the type of place name segment to ensure that a suitable rule base can be selected for place name translation later. In an embodiment of the present application, the place name classifier adopts a neural network model based on deep learning, and establishes a decision boundary between different types of place name data by training a large amount of place name data, so as to map each place name segment to a preset category label, thereby achieving accurate identification of the place name segment type.

[0037] Specifically, the step S32 loads the rule base of the corresponding place name type based on the sequence distribution of the place name fragment type. It should be understood that each rule base contains translation rules formulated for a specific type of place name fragment. For example, for the name of an administrative division, the rule base may contain official standard translation names to ensure the accuracy of the translation; for the name of a natural geographical entity, the rule base may contain translation suggestions based on geographical features and cultural habits to enhance the cultural adaptability and geographical orientation of the translation. By loading a rule base that matches the place name fragment type, the present application can adopt appropriate translation rules and strategies for different types of place name fragments, thereby improving the pertinence and accuracy of the translation. Specifically, the present application classifies and stores different types of place name rule bases by pre-establishing a rule base management system, and establishes an index table in the rule base management system to associate the place name type with the corresponding rule base file path or database table name. After receiving the sequence distribution of the place name fragment type, by traversing the sequence, for each place name fragment type, the corresponding rule base storage location is found according to the index table. For example, if a place name fragment is identified as an "administrative division category", the system finds the storage path of the administrative division category rule library based on the index table, and then loads the rule library into memory through file reading operations or database query operations for subsequent parsing of the place name fragment.

[0038] Specifically, in step S33, based on the rule base of the corresponding place name type, each place name segment in the sequence distribution of the place name segment is parsed to obtain the sequence distribution of the initial candidate translation. Specifically, for each place name segment, a one-by-one matching query is performed in the rule base of the corresponding place name type to find out the possible translation options of each place name segment, that is, the initial candidate translation, so as to obtain the sequence distribution of the initial candidate translation, which is used as a preliminary translation result to lay the foundation for the final place name translation.

[0039] In the above-mentioned method for automatic translation of place names based on a specific rule base, the step S4 performs semantic embedding coding on each initial candidate translated name in the sequence distribution of the initial candidate translated name to obtain a sequence distribution of the semantic embedding coding vector of the initial candidate translated name. It should be understood that considering that translating place name fragments one by one may lead to incoherence and insufficient context adaptability of the overall place name translation result, the present application further introduces natural language processing technology based on deep learning to perform contextual semantic mining and optimization processing on the sequence distribution of the initial candidate translated name to generate a more accurate and semantically fluent place name translation result. In a specific example of the present application, the step S4 includes: using a semantic embedding encoder based on the Bert model to perform semantic embedding coding on each initial candidate translated name in the sequence distribution of the initial candidate translated name to obtain a sequence distribution of the semantic embedding coding vector of the initial candidate translated name. Specifically, this application uses the Bert model, which has excellent performance in natural language processing tasks, as a semantic embedding encoder to perform semantic embedding encoding on each initial candidate translated name, so as to utilize the powerful semantic understanding and representation capabilities of the Bert model obtained by pre-training on a large-scale corpus, capture the deep semantic information contained in each initial candidate translated name, and represent each initial candidate translated name as an embedded vector representation in a high-dimensional semantic space, thereby generating a sequence distribution of the semantic embedding encoding vectors of the initial candidate translated names.

[0040] In the above-mentioned method for automatic translation of place names based on a specific rule base, the step S5 performs contextual association coding based on semantic change perception on the sequence distribution of the semantic embedding coding vector of the initial candidate translated name to obtain the contextual semantic association coding vector of the candidate translated name. It should be understood that in the process of place name translation, there is semantic relevance and coherence between different place name fragments. Therefore, in order to effectively capture the semantic transitions and connections between the initial candidate translated names, the present application adopts a contextual association coding method based on semantic change perception, which captures the semantic transitions and coherence between adjacent translated names by performing semantic association analysis between adjacent translated names on the semantic embedding coding vectors of each initial candidate translated name, thereby dynamically adjusting and optimizing the semantic expression of each initial candidate translated name in the overall place name translation result, so as to enhance the semantic coherence and contextual adaptability of the overall place name translation. Among them, Figure 4 FIG. 5 is a flowchart of sub-step S5 of the method for automatically translating place names based on a specific rule base according to an embodiment of the present application. Figure 4As shown, the step S5 includes the steps of: S51, performing context-neighborhood-aware semantic transfer significance measurement on each initial candidate translated name semantic embedding coding vector in the sequence distribution of the initial candidate translated name semantic embedding coding vector to obtain a sequence distribution of candidate translated name semantic transfer significance factors; S52, based on the sequence distribution of the candidate translated name semantic transfer significance factors, performing feature modulation transfer aggregation coding on the sequence distribution of the initial candidate translated name semantic embedding coding vector to obtain the candidate translated name context semantic association coding vector.

[0041] Figure 5 FIG. 5 is a flowchart of sub-step S51 of the method for automatically translating place names based on a specific rule base according to an embodiment of the present application. Figure 5 As shown, the step S51 includes the steps of: S511, calculating the semantic jump degree of each initial candidate translated name semantic embedding coding vector in the sequence distribution of the initial candidate translated name semantic embedding coding vector to obtain the sequence distribution of the candidate translated name semantic jump degree; S512, calculating the semantic transfer space span of each initial candidate translated name semantic embedding coding vector in the sequence distribution of the initial candidate translated name semantic embedding coding vector to obtain the sequence distribution of the candidate translated name semantic transfer space span; S513, based on the semantic jump degree and semantic transfer space span of each initial candidate translated name semantic embedding coding vector in the sequence distribution of the initial candidate translated name semantic embedding coding vector, calculating the semantic transfer significance factor of each initial candidate translated name semantic embedding coding vector to obtain the sequence distribution of the candidate translated name semantic transfer significance factor.

[0042] More specifically, the step S511 is expressed as follows: ; ; ; in, represents the sequence distribution of the semantic embedding coding vector of the initial candidate translated name, , , and They represent the first, second, and third in the sequence distribution of the semantic embedding coding vector of the initial candidate translated name respectively. and Initial candidate translation semantic embedding encoding vectors, is the number of feature vectors in the sequence distribution of the semantic embedding encoding vector of the initial candidate translated name, Indicates the The first The eigenvalues ​​at the positions, For the The feature scale value of the semantic embedding encoding vector of the initial candidate translation name, For the The semantic strength factor of the semantic embedding encoding vector of the initial candidate translation name, Indicates The semantic strength factor of the semantic embedding encoding vector of the initial candidate translation name, Indicates the The semantic jump degree of the semantic embedding encoding vector of the initial candidate translation name.

[0043] Specifically, in order to keenly capture the fluctuation of semantic association between each initial candidate translated name, the application first measures the drastic degree of semantic change relative to adjacent candidate translated names by calculating the semantic jump degree of the semantic embedding coding vector of each initial candidate translated name. It should be understood that in the context transfer coding process of semantic information of each initial candidate translated name, each initial candidate translated name semantic embedding coding vector performs semantic interaction in sequence according to its order in the original sequence distribution. If the semantic jump degree between a certain initial candidate translated name and the next initial candidate translated name is large, that is, the semantic change is more drastic, it means that there may be a large break or incoherence between the two in semantics. At this point, it is necessary to give a higher semantic attention to the initial candidate translated name for semantic smoothing to ensure the semantic coherence and fluency of the overall place name translation.

[0044] More specifically, the step S512 is expressed as follows: ; in, Indicates the The semantic transfer space span of the semantic embedding encoding vector of the initial candidate translation name, Indicates the and stated The number of eigenvectors between them.

[0045] Specifically, considering that the position of each candidate translation in the overall sequence distribution also has a certain influence on its semantic importance, the present application further calculates the semantic transfer space span of each initial candidate translation semantic embedding coding vector to reveal the position of each candidate translation in the overall sequence distribution, thereby further evaluating its semantic contribution in the overall place name translation result.

[0046] More specifically, the step S513 is expressed by the formula: ; in, and is a preset weight parameter used to balance the influence of semantic jump degree and semantic transfer space span. Indicates the The semantic transfer significance factor of the semantic embedding encoding vector of the initial candidate translated name.

[0047] Specifically, in order to comprehensively consider the two important factors of semantic jump degree and semantic transfer space span, and more accurately measure the importance of each initial candidate translation name semantic embedding coding vector in the context semantic transfer process, the present application further calculates its semantic transfer significance factor based on the semantic jump degree and semantic transfer space span of the semantic embedding coding vector of each initial candidate translation name, so as to focus more on important candidate translation names and reduce attention to secondary information in the subsequent semantic transfer process, thereby making more refined semantic adjustments and optimizations to the semantic feature expressions of each initial candidate translation name.

[0048] Specifically, in step S52, based on the sequence distribution of the semantic transfer significance factor of the candidate translated name, feature modulation transfer aggregation coding is performed on the sequence distribution of the initial candidate translated name semantic embedding coding vector to obtain the candidate translated name context semantic association coding vector. Figure 6 FIG. 5 is a flowchart of sub-step S52 of the method for automatically translating place names based on a specific rule base according to an embodiment of the present application. Figure 6 As shown, the step S52 includes the steps of: S521, inputting the sequence distribution of the candidate translated name semantic transfer significance factors into a gated transfer unit including a softmax normalization function to obtain the sequence distribution of the candidate translated name semantic transfer significance weights; S522, based on the sequence distribution of the candidate translated name semantic transfer significance weights, performing weighted aggregation on the sequence distribution of the initial candidate translated name semantic embedding coding vector to obtain the candidate translated name context semantic association coding vector.

[0049] More specifically, the step S521 is expressed by the formula: ; in, is the normalized exponential function, is the gated mask function, is the gating threshold, For the The candidate translations of are semantically significant.

[0050] Specifically, the present application further introduces a gating mechanism to perform gating screening on the semantic transfer significance factor of each initial candidate translated name semantic embedding coding vector, and generate a sequence distribution of the semantic transfer significance weights of the candidate translated names.

[0051] More specifically, the step S522 is expressed by the formula: ; in, The context semantic association encoding vector representing the candidate translated name.

[0052] Specifically, the sequence distribution of the original initial candidate translation semantic embedding coding vector is weighted and aggregated based on the significant weight of the candidate translation semantics, thereby generating the candidate translation context semantic association coding vector. In this way, the semantic role and importance of each initial candidate translation in the overall place name translation result can be more accurately portrayed, so that the final place name translation result not only accurately conveys the meaning of the original place name, but also is more coherent and fluent in semantics, thereby improving the accuracy and readability of place name translation.

[0053] In the above-mentioned method for automatic translation of place names based on a specific rule base, the step S6 generates a place name translation result based on the candidate translation context semantic association coding vector. In a specific example of the present application, the step S6 includes: after adding a prompt word at the end of the candidate translation context semantic association coding vector, it is input into a translation polisher based on a large language model to obtain the place name translation result. Among them, the prompt word is: "Polish the place name translation based on the context semantic information of the candidate translation". It should be understood that a large language model refers to an artificial intelligence model that has been trained with large-scale text data and has extensive knowledge and complex language understanding capabilities. This type of model can capture the subtle differences in language and make reasonable inferences based on context information. In the application scenario of the present application, the translation polisher uses a pre-trained large language model as the basic framework and fine-tunes the place name translation task so that the model can better understand and handle challenges unique to geographic names, such as the correct conversion of proper nouns and the retention of local characteristics.

[0054] Specifically, the large language model has accumulated rich language knowledge and translation experience by training on large-scale corpus. Therefore, after receiving the context semantic association coding vector of the candidate translated name containing the prompt word, the large language model can deeply understand and analyze the context semantic association coding vector of the candidate translated name according to the instruction of the prompt word, comprehensively consider the semantic information and context association of the candidate translated name, and use its powerful language generation ability and translation experience to generate accurate and fluent place name translation results.

[0055] In a preferred example of the present application, since each initial candidate translated name semantic embedding coding vector in the sequence distribution of the initial candidate translated name semantic embedding coding vector respectively represents the semantic embedding coding features of the initial candidate translated name of the corresponding place name fragment, when performing context semantic encoding based on feature sequence message transmission of feature jump degree, the difference in the attribute of the embedded coding semantic feature population in the local semantic space will cause fairness difference in the feature jump degree level, thereby affecting the context semantic association distribution transmission inclusiveness of the candidate translated name context semantic association coding vector in the global semantic space, and reducing the semantic quality of the place name translation result obtained by the translation polisher based on the large language name as input.

[0056] Therefore, in this preferred example, after adding the prompt word to the tail of the candidate translated name context semantic association encoding vector, when inputting the translation polisher based on the large language name, the candidate translated name context semantic association encoding vector is first optimized, and the specific optimization process is: Determine the number of super-distributed feature values ​​in the candidate translated name context semantic association encoding vector whose difference from the feature mean is greater than a threshold, and respectively calculate the reciprocal of the logarithm of the number of super-distributed feature values ​​with base two and the exponential value of the reciprocal of the number of super-distributed feature values ​​with a natural constant as the base to obtain the first candidate translated name context semantic association nonlinear trajectory representation value The nonlinear trajectory representation value associated with the contextual semantics of the second candidate translation : ; ; ; in, represents the context semantic association encoding vector of the candidate translation, The first eigenvalues, represents the mean value of all feature values ​​of the candidate translated name context semantic association encoding vector, Representation calculation The number of is the number of super-distributed eigenvalues ​​in the candidate translated name context semantic association encoding vector, and is the threshold hyperparameter; Calculate the hyperbolic sine function value of the sum of squares of all eigenvalues ​​of the candidate translated name context semantic association encoding vector: ; in, represents the length of the candidate translation context semantic association encoding vector, represents the hyperbolic sine function, Represents the value of the hyperbolic sine function; The exponential value of the hyperbolic sine function value with the natural constant as the base is calculated and then divided by the square of the length of the candidate translated name context semantic association encoding vector to obtain the candidate translated name context semantic association potential manifold value ,in, represents a natural constant, Represents the potential manifold value of the contextual semantic association of candidate translations; Calculate the potential manifold value of the contextual semantic association of the candidate translation The nonlinear trajectory representation value of the contextual semantic association with respect to the first candidate translation The nonlinear trajectory representation value of the contextual semantic association with the second candidate translation The integrated value of the covariance of the contextual semantic association of the candidate translated name: ; in, Indicates the integrated value of the covariance of the contextual semantic association of the candidate translation; Calculate the power function feature vector of the candidate translated name context semantic association encoding vector with the candidate translated name context semantic association covariance integration value as the inverse, and calculate the autocorrelation matrix of the power function feature vector to obtain the candidate translated name context semantic association adaptive representation matrix, that is: ; in, represents a power function feature vector whose inverse is the integrated value of the candidate translated name context semantic association covariance for calculating the candidate translated name context semantic association encoding vector, represents matrix multiplication, represents the transpose of a vector, An adaptive representation matrix representing the contextual semantic association of candidate translations; The candidate translated name context semantic association encoding vector is matrix-multiplied with the candidate translated name context semantic association adaptive representation matrix to obtain an optimized candidate translated name context semantic association encoding vector, that is: ; Here, the candidate translation context semantic association encoding vector is a row vector, Represents the optimized context semantic association encoding vector of candidate translations.

[0057] Specifically, for the dynamic mode cluster of the context semantic association encoding vector of the candidate translated name under the constraint condition of random field trajectory generation, the cross-modal positioning energy consumption fluctuation phenomenon mainly originates from the global action threshold that exceeds the adjacent scope. By implementing the dynamic manifold hidden variable analysis method based on covariance tensor decomposition, the implicit heterogeneous association map in its feature space is analyzed, and by constructing an elastic manifold hidden variable framework with dimensional elasticity, the dynamic coupling characterization protocol between dynamic interactive time series projection parameters is reestablished. Therefore, through this inverse modeling scheme of the energy state transition path, discrete phase change modeling under global parameter association is effectively realized, and the context semantic association distribution transmission inclusiveness of the candidate translated name context semantic association encoding vector in the global semantic space is significantly improved. In this way, the accuracy of the place name translation result obtained by the translation polisher based on the large language name is improved.

[0058] In order to maintain the original cultural characteristics and regional orientation of place names, the translation polisher has a built-in evaluation system to measure whether the candidate translations meet the established standards. This evaluation system takes into account a variety of factors, such as phonetic beauty, consistency of literal meaning, local idioms, etc., to ensure that the output results are both scientific and reasonable and resonate. In addition, the system introduces a user feedback mechanism, which allows the performance of the model to be continuously improved through correction suggestions in actual applications, thereby achieving continuous optimization. This dynamic adjustment process not only improves the performance of the model, but also accumulates valuable experience for future place name translation tasks.

[0059] During this process, special attention should be paid to protecting user privacy and data security. All transmitted data adopts encryption protocol to ensure that the information will not be intercepted or tampered with during the transmission process. At the same time, it is necessary to strictly abide by the relevant laws and regulations on the processing of personal information to ensure that the rights and interests of users are not infringed. In addition, in order to improve the transparency and trust of the system, the system also provides detailed log records and progress reports to keep users informed of the status and progress of the task at any time.

[0060] In summary, the method for automatic translation of place names based on a specific rule base according to the embodiment of the present application is explained, which first splits the place name to be translated input by the user to obtain multiple place name fragments, and loads the corresponding rule base based on the type of each place name fragment to parse each place name fragment to obtain multiple initial candidate translated names, and then introduces natural language processing technology based on deep learning to perform semantic analysis and contextual semantic association encoding on each initial candidate translated name to obtain the global contextual semantic association features of the candidate translated name, and with the help of a large language model, the place name is translated based on the contextual semantic information of the candidate translated name to generate a place name translation result. By combining a specific rule base and semantic parsing technology, this method can comprehensively consider the language characteristics, cultural background and contextual context of the place name, thereby improving the accuracy of place name translation.

[0061] The basic principle of the present invention is described above in conjunction with specific embodiments. However, it should be pointed out that the advantages, strengths, effects, etc. mentioned in the present invention are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. must be possessed by each embodiment of the present invention. In addition, the specific details of the above embodiments are only for the purpose of illustration and facilitation of understanding, rather than limitation, and the above details do not limit the present invention to being implemented by adopting the above specific details.

[0062] In the above embodiments, the description of each embodiment has its own emphasis. For the parts that are not described or recorded in detail in a certain embodiment, please refer to the relevant description of other embodiments. In the several embodiments provided by the present invention, it should be understood that the disclosed system and method can be implemented in other ways. For example, the system embodiment described above is only schematic. For example, the unit division is only a logical function division, and there may be other division methods in actual implementation. The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0063] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference to a figure in a claim should not be considered as limiting the claim to which it relates.

[0064] In addition, it is obvious that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. The multiple units stated in the system claims can also be implemented by one unit through software or hardware.

[0065] Finally, it should be noted that the above description has been given for the purpose of illustration and description. In addition, the above embodiments are only used to illustrate the technical solution of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A method for automatically translating place names based on a specific rule base, characterized in that: include: Obtaining the place name to be translated input by the user; Splitting the place name to be translated to obtain a sequence distribution of place name fragments; Based on the type of each place name segment in the sequence distribution of the place name segment, the each place name segment is parsed based on a specific rule base to obtain a sequence distribution of initial candidate translated names; Performing semantic embedding coding on each initial candidate translated name in the sequence distribution of the initial candidate translated names to obtain a sequence distribution of semantic embedding coding vectors of the initial candidate translated names; Performing context-association coding based on semantic change perception on the sequence distribution of the initial candidate translated name semantic embedding coding vector to obtain a candidate translated name context semantic association coding vector; Based on the candidate translated name context semantic association coding vector, a place name translation result is generated.

2. The method for automatically translating place names based on a specific rule base according to claim 1 is characterized in that: Based on the type of each place name segment in the sequence distribution of the place name segment, the each place name segment is parsed based on a specific rule base to obtain the sequence distribution of the initial candidate translated name, including: Identifying the type of each place name segment in the sequence distribution of the place name segments based on the place name classifier to obtain a sequence distribution of the place name segment types; Based on the sequence distribution of the place name segment type, loading a rule library of the corresponding place name type; Based on the rule base of the corresponding place name type, each place name segment in the sequence distribution of the place name segments is parsed to obtain the sequence distribution of the initial candidate translated names.

3. The method for automatically translating place names based on a specific rule base according to claim 2 is characterized in that: Performing semantic embedding coding on each initial candidate translated name in the sequence distribution of the initial candidate translated names to obtain a sequence distribution of semantic embedding coding vectors of the initial candidate translated names, including: A semantic embedding encoder based on the Bert model is used to perform semantic embedding encoding on each initial candidate translated name in the sequence distribution of the initial candidate translated names to obtain a sequence distribution of the semantic embedding encoding vectors of the initial candidate translated names.

4. The method for automatically translating place names based on a specific rule base according to claim 3 is characterized in that: The sequence distribution of the initial candidate translated name semantic embedding coding vector is subjected to contextual association coding based on semantic change perception to obtain a candidate translated name contextual semantic association coding vector, including: Performing a context-neighborhood-aware semantic transfer significance measurement on each initial candidate translated name semantic embedding coding vector in the sequence distribution of the initial candidate translated name semantic embedding coding vector to obtain a sequence distribution of candidate translated name semantic transfer significance factors; Based on the sequence distribution of the semantic transfer significance factors of the candidate translated names, feature modulation transfer aggregation coding is performed on the sequence distribution of the initial candidate translated name semantic embedding coding vector to obtain the candidate translated name context semantic association coding vector.

5. The method for automatically translating place names based on a specific rule base according to claim 4 is characterized in that: The method of performing a context-neighborhood-aware semantic transfer significance measurement on each initial candidate translated name semantic embedding coding vector in the sequence distribution of the initial candidate translated name semantic embedding coding vector to obtain a sequence distribution of candidate translated name semantic transfer significance factors includes: Calculating the semantic jump degree of each initial candidate translated name semantic embedding coding vector in the sequence distribution of the initial candidate translated name semantic embedding coding vector to obtain the sequence distribution of the candidate translated name semantic jump degree; Calculating the semantic transfer space span of each initial candidate translated name semantic embedding coding vector in the sequence distribution of the initial candidate translated name semantic embedding coding vector to obtain the sequence distribution of the candidate translated name semantic transfer space span; Based on the semantic jump degree and semantic transfer space span of each initial candidate translated name semantic embedding coding vector in the sequence distribution of the initial candidate translated name semantic embedding coding vector, the semantic transfer significance factor of each initial candidate translated name semantic embedding coding vector is calculated to obtain the sequence distribution of the candidate translated name semantic transfer significance factor.

6. The method for automatically translating place names based on a specific rule base according to claim 5 is characterized in that: Based on the sequence distribution of the semantic transfer significance factor of the candidate translated name, feature modulation transfer aggregation coding is performed on the sequence distribution of the initial candidate translated name semantic embedding coding vector to obtain the candidate translated name context semantic association coding vector, including: Inputting the sequence distribution of the semantic transfer significance factors of the candidate translated names into a gated transfer unit including a softmax normalization function to obtain the sequence distribution of the semantic transfer significance weights of the candidate translated names; Based on the sequence distribution of the candidate translated name semantically significant weights, weighted aggregation is performed on the sequence distribution of the initial candidate translated name semantic embedding coding vectors to obtain the candidate translated name context semantic association coding vector.

7. The method for automatically translating place names based on a specific rule base according to claim 6 is characterized in that: Based on the candidate translated name context semantic association encoding vector, generating a place name translation result, including: After adding a prompt word to the tail of the candidate translated name context semantic association encoding vector, it is input into a translation polisher based on a large language model to obtain the place name translation result.

8. The method for automatically translating place names based on a specific rule base according to claim 7 is characterized in that: The prompt word is: "polish the place name translation based on the contextual semantic information of the candidate translated names."

Images