Chinese webpage internationalization adaptation method and device based on large model and medium

Abstract

The invention provides a Chinese webpage internationalization adaptation method and device based on a large model and a medium, and belongs to the technical field of networks. The Chinese webpage internationalization adaptation method comprises the following steps: analyzing a Chinese webpage by using a natural language model and a webpage analysis algorithm to obtain Chinese text information, multimedia resource information and webpage architecture information; performing intelligent translation and domain term adaptation on the Chinese text information according to the large semantic understanding model in combination with a multi-language corpus and a domain term library to obtain internationalized text information; performing culture adaptation on the internationalized text information, the multimedia resource information and the webpage architecture information according to a culture environment suitable for the internationalized text information to obtain an internationalized resource file; and loading the internationalized resource file according to the current language environment to obtain an internationalized webpage in the current language environment. According to the method, the problems that existing webpage internationalization tools generally lack intelligence, cannot accurately understand semantics and contexts of Chinese webpages and are difficult to adapt to internationalization requirements can be solved.

Description

Technical Field

[0001] This application belongs to the technical field of digital networks, and particularly relates to a method, device, and medium for internationalizing Chinese web pages based on large models. Background Art

[0002] With the development of digital network technology, the Internet shows a general trend of globalization and internationalization, and the internationalization demand for Chinese web pages also shows an increasing trend, which requires the internationalization adaptation of Chinese web pages. The traditional internationalization adaptation of Chinese web pages generally adopts the method of manually translating content. To adapt to regional characteristics, corresponding web page version modes need to be developed separately for different languages and regions. The traditional method has many drawbacks: First, the manual translation process not only consumes a large amount of time and labor costs, but also is extremely prone to translation errors, resulting in inaccurate information transmission; Second, developing web page versions separately for different languages and regions greatly increases the development cost and maintenance cost, and at the same time, it is difficult to quickly respond to market changes and the diverse needs of users.

[0003] To solve the above drawbacks, automated web page internationalization tools have been developed in the prior art. For example, the existing patent (CN115796190A) discloses a front-end internationalization multi-language conversion method based on vue and webpack. Specifically, for an application project built with vue and webpack, when it is used for the first time, project initialization configuration is performed through the project configuration module; then the Chinese in the project is extracted and stored in the language package storage directory configured in the set file; then the translation module is used to translate the files in the directory, automatically translate and generate language packages, and automatically convert Chinese into language variables during the project operation or packaging process. Specifically, in the translation module, a translation API is called to translate Chinese character strings until the language translation is completed. Most of the existing translation APIs are connected to natural language models. For example, by using the associated information of the text to be translated, Chinese texts are searched to construct a corpus; then the term corpus in the corpus and the text representation vectors are input into a recurrent neural network model with a pre-trained multi-head attention mechanism to obtain the target translation text.

[0004] However, the above-mentioned automated web page internationalization tools generally lack intelligence, cannot accurately understand the semantics and context of Chinese web pages, and are difficult to meet the internationalization requirements. Problems such as content disorder and format mismatch frequently occur during the adaptation process, and it is difficult to meet the requirements of efficient and accurate internationalization adaptation. Summary of the Invention

[0005] This application aims to provide an intelligent and efficient internationalization adaptation solution for Chinese web pages based on large models. By using a semantic understanding large model and combining a preset multilingual corpus and domain term library, it intelligently translates Chinese text information and adapts domain terms to obtain internationalized text information. Then, it performs cultural adaptation on the internationalized text information, multimedia resource information, and web page architecture information corresponding to the Chinese web page to obtain an internationalized resource file. Finally, when the web page needs to run, loading this internationalized resource file can solve the technical problems that existing web page internationalization tools generally lack intelligence and are difficult to accurately understand the semantics and context of Chinese web pages.

[0006] According to the first aspect of this application, an embodiment of this application provides a method for internationalization adaptation of Chinese web pages based on large models, including: Using a natural language model and a web page parsing algorithm to parse a Chinese web page to obtain the Chinese text information, multimedia resource information, and web page architecture information of the Chinese web page; According to the semantic understanding large model, combining a preset multilingual corpus and domain term library, performing intelligent translation and domain term adaptation on the Chinese text information to obtain internationalized text information; According to the cultural environment applicable to the internationalized text information, performing cultural adaptation on the internationalized text information, multimedia resource information, and web page architecture information to obtain an internationalized resource file; When the web page needs to run, loading the internationalized resource file according to the current language environment to obtain an internationalized web page corresponding to the current language environment.

[0007] Preferably, in the above method for internationalization adaptation of Chinese web pages, the step of using a natural language model and a web page parsing algorithm to parse a Chinese web page to obtain the Chinese text information, multimedia resource information, and web page architecture information of the Chinese web page includes: Using a natural language model to parse a Chinese web page to obtain the Chinese text information corresponding to the Chinese web page; Using a natural language model to split the Chinese text information into lexical and sentence forms, and analyzing to obtain the grammatical structure and semantic information in lexical and sentence forms; Using a web page parsing algorithm to parse a Chinese web page to obtain the multimedia resource information and web page architecture information corresponding to the Chinese web page; Classifying and identifying the multimedia resource information.

[0008] Preferably, in the above method for internationalization adaptation of Chinese web pages, before the step of performing intelligent translation and domain term adaptation on the Chinese text information according to the semantic understanding large model and combining a preset multilingual corpus and domain term library to obtain internationalized text information, it further includes: Refining the keywords in the Chinese text information as key values; Replace keywords in the Chinese text information with key values and add multilingual tags to the key values to obtain new Chinese text information; Generate a Chinese resource file containing the new Chinese text information; Input the Chinese resource file into the semantic understanding large model.

[0009] Preferably, in the above Chinese web page internationalization adaptation method, according to the semantic understanding large model, combined with a preset multilingual corpus and domain term library, the steps of performing intelligent translation and domain term adaptation on the Chinese text information to obtain internationalized text information include: Construct a general semantic space using the multilingual corpus and train the semantic understanding large model using the multilingual corpus so that the semantic understanding large model learns the general semantic rules corresponding to the multilingual corpus; Input the Chinese text information into the semantic understanding large model to learn the grammatical structure and semantic information corresponding to the Chinese text information, where the Chinese text information includes keywords; Control the semantic understanding large model to learn the semantic association relationship between the keyword and the context and capture the semantic role corresponding to the keyword in combination with the grammatical structure and semantic information; Control the semantic understanding large model to map the keyword to the general word vector in the semantic space using the general semantic rules according to the semantic association relationship and semantic role; And, construct a professional semantic space using the domain term library and train the semantic understanding large model using the domain term library so that the semantic understanding large model learns the professional semantic rules corresponding to the domain term library; Control the semantic understanding large model to map the keyword containing professional domain terms to the professional word vector in the professional semantic rules according to the professional semantic rules; According to the mapping relationship between Chinese and other language texts in the multilingual corpus, map the general word vector and the professional word vector to foreign language texts respectively to obtain internationalized text information.

[0010] Preferably, in the above Chinese web page internationalization adaptation method, the steps of performing cultural adaptation on the internationalized text information, multimedia resource information, and web page architecture information according to the cultural environment applicable to the internationalized text information to obtain an internationalized resource file include: Obtain the cultural environments of different countries and regions to form a cultural feature library, and use the cultural features in the cultural feature library to train a multimodal analysis model to construct a cultural semantic space; where the multimodal deep analysis model is a two-stream heterogeneous deep network architecture; Use the text processing stream of the two-stream heterogeneous deep network architecture to parse the internationalized text information according to the cultural semantic space and extract the culturally sensitive text features in the internationalized text information; A visual processing stream using a dual-stream heterogeneous deep network architecture, which uses an object detection algorithm combined with a multi-scale attention mechanism to respectively detect the feature associations in multimedia resource information and web page architecture information, and obtains the culture-sensitive visual features corresponding to the multimedia resource information or the web page architecture information; Replace the culture-sensitive text features according to the cultural semantic space and adjust the display of the culture-sensitive visual features.

[0011] Preferably, after the step of performing cultural adaptation on internationalized text information, multimedia resource information, and web page architecture information to obtain an internationalized resource file in the above Chinese web page internationalization adaptation method, the following steps are further included: Render the web page using common rendering methods by simulating multiple target resolutions respectively; Switch the display effects of each internationalized resource file on the web page in turn; When there are problems with the display effect of the internationalized resource file on the web page, adjust the web page architecture information.

[0012] Preferably, in the above Chinese web page internationalization adaptation method, the step of loading an internationalized resource file according to the current language environment to obtain an internationalized web page corresponding to the current language environment includes: When the web page needs to be run, load the HTML script corresponding to the current web page, where the HTML script includes the current language environment; Use the HTML script to parse the internationalized text information in the internationalized resource file; Load the internationalized resource file according to the current language environment to obtain an internationalized web page in the current language environment.

[0013] According to the second aspect of the present application, the present application also provides a Chinese web page internationalization adaptation device based on a large model, including: A web page parsing module for parsing a Chinese web page using a natural language model and a web page parsing algorithm to obtain Chinese text information, multimedia resource information, and web page architecture information of the Chinese web page; An information adaptation module for performing intelligent translation and domain term adaptation on the Chinese text information according to a semantic understanding large model in combination with a preset multilingual corpus and a domain term library to obtain internationalized text information; A cultural adaptation module for performing cultural adaptation on the internationalized text information, multimedia resource information, and web page architecture information according to the cultural environment applicable to the internationalized text information to obtain an internationalized resource file; A file loading module for loading an internationalized resource file according to the current language environment to obtain an internationalized web page corresponding to the current language environment when the web page needs to be run.

[0014] Preferably, in the above-mentioned Chinese web page internationalization adaptation device based on a large model, the web page parsing module is further configured to parse the Chinese web page using a natural language model to obtain the Chinese text information corresponding to the Chinese web page; split the Chinese text information into words and sentences using the natural language model, and analyze the grammatical structure and semantic information in the form of words and sentences; parse the Chinese web page using a web page parsing algorithm to obtain the multimedia resource information and web page architecture information corresponding to the Chinese web page; classify and label the multimedia resource information.

[0015] Preferably, the above-mentioned Chinese web page internationalization adaptation device based on a large model further includes: a Chinese resource file generation module, configured to extract the keywords in the Chinese text information as key values; replace the keywords with the key values in the Chinese text information, and add multilingual tags to the key values to obtain new Chinese text information; generate a Chinese resource file containing the new Chinese text information; and input the Chinese resource file into the semantic understanding large model.

[0016] Preferably, the above-mentioned information adaptation module is specifically configured to construct a general semantic space using a multilingual corpus and train a semantic understanding large model using the multilingual corpus, so that the semantic understanding large model learns the general semantic rules corresponding to the multilingual corpus; input the Chinese text information into the semantic understanding large model to learn the grammatical structure and semantic information corresponding to the Chinese text information, where the Chinese text information includes keywords; control the semantic understanding large model to learn the semantic association relationship between the keywords and the context and capture the semantic roles corresponding to the keywords in combination with the grammatical structure and semantic information; control the semantic understanding large model to map the keywords to the general word vectors in the semantic space using the general semantic rules according to the semantic association relationship and semantic roles; and construct a professional semantic space using a domain term library and train the semantic understanding large model using the domain term library, so that the semantic understanding large model learns the professional semantic rules corresponding to the domain term library; control the semantic understanding large model to map the keywords containing professional domain terms to the professional word vectors in the professional semantic rules according to the professional semantic rules; and map the general word vectors and professional word vectors to foreign language texts respectively according to the mapping relationship between Chinese and other language texts in the multilingual corpus to obtain internationalized text information.

[0017] Preferably, in the above-mentioned Chinese web page internationalization adaptation device based on a large model, the culture adaptation module is specifically configured to obtain the cultural environments of different countries and regions, form a cultural feature library, use the cultural features in the cultural feature library to train a multi-modal analysis model, and construct a cultural semantic space; among them, the multi-modal deep analysis model is a two-stream heterogeneous deep network architecture; use the text processing stream of the two-stream heterogeneous deep network architecture to parse the internationalized text information according to the cultural semantic space, and extract the culturally sensitive text features in the internationalized text information; use the visual processing stream of the two-stream heterogeneous deep network architecture, and use the object detection algorithm combined with the multi-scale attention mechanism to respectively detect the feature associations in the multimedia resource information and the web page architecture information, and obtain the culturally sensitive visual features corresponding to the multimedia resource information or the web page architecture information; replace the culturally sensitive text features according to the cultural semantic space, and adjust the display of the culturally sensitive visual features.

[0018] Preferably, in the above-mentioned Chinese web page internationalization adaptation device based on a large model, the file loading module is specifically configured to load the HTML script corresponding to the current web page when the web page needs to be run, where the HTML script includes the current language environment; use the HTML script to parse the internationalized text information in the internationalized resource file; load the internationalized resource file according to the current language environment to obtain the internationalized web page in the current language environment.

[0019] According to the third aspect of the present application, the present application also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the Chinese web page internationalization adaptation method provided by any of the above technical solutions.

[0020] According to the fourth aspect of the present application, the present application also provides a computer storage medium, on which computer executable instructions are stored. When the computer program is executed by the processor, it implements the Chinese web page internationalization adaptation method provided by any of the above technical solutions.

[0021] The technical solution of the present application has at least the following technical effects: The Chinese web page internationalization adaptation solution provided by this application first uses a natural language model and a web page parsing algorithm to parse Chinese web pages, so as to obtain the Chinese text information, multimedia resource information, and web page architecture information of the Chinese web pages. Then, a semantic understanding large model is used to perform intelligent translation and domain term adaptation on the Chinese text information in combination with a preset multilingual corpus and a domain term library, so as to obtain internationalized text information. Because the semantic understanding large model is used in combination with the preset multilingual corpus and the domain term library to perform intelligent translation and domain term adaptation on the Chinese text information after the above natural language parsing, the semantics and context in the Chinese text information can be accurately understood, and the cultural differences in different countries and regions can be fully considered, and the professional terms in related fields can be accurately adapted, so as to reduce the problems of content disorder and format mismatch in the adaptation process. In addition, cultural adaptation is performed on the internationalized text information, multimedia resource information, and web page architecture information in combination with the cultural environment to which the internationalized text needs to be applied, so as to meet the international cultural needs. Finally, when the web page needs to be run, the internationalized resource file is loaded according to the current language environment, and the internationalized web page corresponding to the current language environment can be obtained. Through the above method, the problems in the prior art that the web page internationalization tools generally lack intelligence, it is difficult to accurately understand the semantics and context of Chinese web pages, and it is difficult to meet the internationalization requirements can be solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The drawings described herein are used to provide a further understanding of this application, and constitute a part of this application. The illustrative embodiments of this application and their descriptions are used to explain this application and do not constitute an improper limitation of this application. In the drawings: Figure 1 is a schematic flowchart of a method for internationalizing and adapting Chinese web pages based on a large model provided by an embodiment of this application; Figure 2 is Figure 1 a schematic flowchart of a method for parsing a Chinese web page provided by the illustrated embodiment; Figure 3 is a schematic flowchart of a method for marking keywords in Chinese text information provided by an embodiment of this application; Figure 4 is Figure 1 a schematic flowchart of a method for intelligent translation and domain term adaptation of Chinese text information provided by the illustrated embodiment; Figure 5 is Figure 1 a schematic flowchart of a method for cultural adaptation of an internationalized resource file provided by the illustrated embodiment; Figure 6 is a schematic flowchart of a method for formatting and adapting a web page provided by an embodiment of this application; Figure 7 is Figure 1Schematic flowchart of a method for loading an internationalization resource file provided by the illustrated embodiment; Figure 8 Schematic structural diagram of a Chinese web page internationalization adaptation device based on a large model provided by an embodiment of the present application; Figure 9 Schematic structural diagram of an electronic device provided by an embodiment of the present application. Detailed implementation manners

[0023] To more clearly illustrate the overall concept of the present application, the following will be described in detail by way of examples in conjunction with the accompanying drawings of the specification.

[0024] In the following description, many specific details are set forth in order to fully understand the present application. However, the present application may also be implemented in other ways different from those described herein. Therefore, the scope of protection of the present application is not limited by the specific embodiments disclosed below. It should be noted that, without conflict, the embodiments of the present application and the features in each embodiment may be combined with each other.

[0025] In the present application, unless otherwise clearly defined and limited, the first feature being "on" or "under" the second feature may be that the first and second features are in direct contact, or the first and second features are in indirect contact through an intermediate medium. In the description of this specification, the description with reference to terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0026] The prior art has the following defects: Existing automated web page internationalization tools generally lack intelligence, are unable to accurately understand the semantics and context of Chinese web pages, and are difficult to meet internationalization requirements. Problems such as content disorder and format mismatch frequently occur during the adaptation process, making it difficult to meet the requirements of efficient and accurate internationalization adaptation.

[0027] To solve the above problems, the following embodiments of the present invention provide a Chinese web page internationalization adaptation solution based on a large model, which can solve the problems that the prior art is difficult to accurately understand the semantics and context of Chinese web pages, difficult to meet internationalization requirements, and prone to disorder and format mismatch. Thus, the purpose of quickly and accurately adapting Chinese web pages to various internationalization environments, effectively reducing development and maintenance costs, and significantly improving the adaptation efficiency and quality is achieved.

[0028] To achieve the above object, refer to Figure 1, this application provides a schematic flowchart of a method for internationalizing and adapting Chinese web pages based on large models. As Figure 1 shown, the method for internationalizing and adapting Chinese web pages based on large models includes: S110: Use a natural language model and a web page parsing algorithm to parse a Chinese web page to obtain the Chinese text information, multimedia resource information, and web page architecture information of the Chinese web page.

[0029] Natural language processing (NLP), as a core field of artificial intelligence, the natural language processing model here can select the word embedding model Word2Vec, the generative pre-training module GPT, and the dynamic word vector model ELMo, etc. The web page parsing algorithm can select the Python crawler algorithm. The embodiments of this application combine a natural language model and a web page parsing algorithm to parse a Chinese web page, which can perform text analysis and information extraction on the Chinese web page, so as to obtain a large amount of valuable information from the web page. Specifically, in the embodiments of this application, the Chinese text information, multimedia resource information, and web page architecture information of the Chinese web page can be obtained. Furthermore, intelligent translation is performed on the Chinese text information, and adaptation of the multimedia resource information and web page architecture information to the professional field and culture is carried out, so as to obtain an internationalized web page suitable for the needs of the destination.

[0030] Specifically, as a preferred embodiment, as Figure 2 shown, the above step S110: Use a natural language model and a web page parsing algorithm to parse a Chinese web page to obtain the Chinese text information, multimedia resource information, and web page architecture information of the Chinese web page, specifically including: S111: Use a natural language model to parse a Chinese web page to obtain the Chinese text information corresponding to the Chinese web page.

[0031] S112: Use a natural language model to split the Chinese text information into vocabulary and sentence forms, and analyze to obtain the grammatical structure and semantic information in the form of vocabulary and sentences.

[0032] S113: Use a web page parsing algorithm to parse a Chinese web page to obtain the multimedia resource information and web page architecture information corresponding to the Chinese web page.

[0033] S114: Classify and label the multimedia resource information.

[0034] The technical solution provided by the embodiments of this application can deeply parse a Chinese web page, comprehensively use natural language processing technology and web page parsing algorithms to comprehensively and deeply parse the Chinese web page, so as to obtain the Chinese text information mainly based on the text content of the web page, as well as multimedia resource information such as pictures, audio, and video, and obtain web page architecture information including the HTML structure and CSS styles of the Chinese web page. The web page architecture information obtained here can be adaptively adjusted in format according to the display requirements of the destination.

[0035] Among them, for Chinese text information, it is necessary to split the Chinese text in the Chinese text information into words and sentence forms, and deeply analyze its grammatical structure and semantic information to facilitate the subsequent accurate understanding of the semantics and context of the Chinese text information by the large model; at the same time, classify and identify the multimedia resources on the web page to lay a foundation for subsequent processing of professional field terms and cultural adaptation.

[0036] In addition, as a preferred embodiment, as Figure 3 shown, after the above Chinese web page internationalization adaptation method, in step 110: using a natural language model and a web page parsing algorithm to parse the Chinese web page to obtain the Chinese text information, multimedia resource information, and web page architecture information of the Chinese web page, it further includes: S201: Extract the keywords in the Chinese text information as the key value; S202: Use the key value to replace the keyword in the Chinese text information and add a multilingual tag to the key value to obtain new Chinese text information; S203: Generate a Chinese resource file containing the new Chinese text information; S204: Input the Chinese resource file into the semantic understanding large model.

[0037] The technical solution provided by the embodiment of the present application needs to process the Chinese text after obtaining the Chinese text information. Specifically, the Chinese text is refined and summarized into keywords as the key value, the Chinese text in the HTML code is replaced with the corresponding key value, and multilingual tags __i18n__ are added at its beginning and end to facilitate subsequent efficient replacement and mapping of keywords. After processing the Chinese text, a Chinese language resource file is generated. Specifically, a Chinese resource file is automatically generated based on the recognized and parsed Chinese text and key value, which can be named zh_CHS.json; if the Chinese resource file already exists, the new content is appended to the file and automatically sorted in alphabetical order of the key. In this way, subsequent use of the large model can perform replacement, mapping, and translation in alphabetical order of the key, thereby improving data processing efficiency and avoiding garbled characters.

[0038] Figure 1 The technical solution provided by the shown embodiment, after the step of using a natural language model and a web page parsing algorithm to parse the Chinese web page to obtain the Chinese text information, multimedia resource information, and web page architecture information of the Chinese web page, further includes: S120: Based on the semantic understanding large model, combined with the preset multilingual corpus and domain term library, perform intelligent translation and domain term adaptation on the Chinese text information to obtain internationalized text information. From the above content, it can be seen that after generating a Chinese resource file containing new Chinese text information in this application, the Chinese resource file is input into the semantic understanding large model, which can directly parse the Chinese resource file, specifically perform intelligent translation and domain term adaptation on the Chinese text information, and can be understood as directly performing intelligent translation and domain term adaptation on the new Chinese text information contained in the Chinese resource file.

[0039] In the embodiment of this application, the intelligent ability (i.e., the large model) is called to translate the Chinese text information contained in the Chinese resource file, and at the same time, domain adaptation and cultural adaptation work are carried out to generate resource files in other languages, such as English: en.json and / or Traditional Chinese: zh_CHT.json, etc.

[0040] Specifically, as a preferred embodiment, as Figure 4 shown, in the above Chinese web page internationalization adaptation method, step S120: The step of performing intelligent translation and domain term adaptation on the Chinese text information according to the semantic understanding large model, combined with the preset multilingual corpus and domain term library, to obtain internationalized text information specifically includes: S121: Construct a general semantic space using the multilingual corpus, and train the semantic understanding large model using the multilingual corpus, so that the semantic understanding large model learns the general semantic rules corresponding to the multilingual corpus.

[0041] S122: Input the Chinese text information into the semantic understanding large model to learn the grammatical structure and semantic information corresponding to the Chinese text information, where the Chinese text information includes keywords.

[0042] S123: Control the semantic understanding large model, combined with the grammatical structure and semantic information, to learn the semantic association relationship between the keywords and the context, and capture the semantic roles corresponding to the keywords.

[0043] S124: Control the semantic understanding large model, according to the semantic association relationship and semantic roles, use the general semantic rules to map the keywords to the general word vectors in the semantic space.

[0044] First, use a multilingual corpus combined with a semantic understanding large model to learn general semantic rules. After the learning and training are completed, input the above Chinese text information into the semantic understanding large model, and it can learn the grammatical structure and semantic information corresponding to the Chinese text information. The grammatical structure and semantic information can be parsed by the natural language model in the previous step. Here, use the semantic understanding large model to learn the grammatical structure and semantic information, and then it can learn the semantic association relationship between the above keywords and the context, and capture the semantic roles of the keywords (such as agent, patient, time, and location, etc.). Specifically, in the sentence "She took a walk in the park yesterday", the model can label "yesterday" as an adverbial of time and "the park" as an adverbial of place, thus constructing a complete context. Then the semantic understanding large model can use the general semantic rules to learn the semantic association relationship and semantic roles of the above keywords, so as to accurately understand the semantics and context of the keywords in the Chinese text, and map the colloquial or informal keywords to formal general word vectors in the semantic space. Such general word vectors can accurately express the semantic and context characteristics of the keywords in the Chinese text, thus realizing the accurate translation of the word vectors.

[0045] And, S125: Construct a professional semantic space using a domain term library, and train the semantic understanding large model using the domain term library, so that the semantic understanding large model learns the professional semantic rules corresponding to the domain term library.

[0046] S126: Control the semantic understanding large model to map the keywords containing domain-specific terms to the professional word vectors in the professional semantic rules according to the professional semantic rules. The method steps of mapping the keywords containing domain-specific terms to the professional word vectors in the professional semantic rules according to the professional semantic rules are similar to the method of keyword mapping using general semantic rules above, and will not be elaborated here.

[0047] S127: According to the mapping relationship between Chinese and other language texts in the multilingual corpus, map the general word vectors and professional word vectors to foreign language texts respectively to obtain internationalized text information.

[0048] For terms in the professional field, the embodiments of the present application can also use the semantic understanding large model to learn the professional semantic rules corresponding to the domain term library, so as to map the keywords containing domain-specific terms to the professional word vectors in the professional semantic rules according to the professional semantic rules, accurately understand the meaning of the Chinese text, and accurately translate the professional terms.

[0049] In summary, the technical solution provided by the embodiments of the present application combines a multilingual corpus and a large semantic understanding model, accurately understands the meaning of Chinese text information based on the context and context of keywords in Chinese text information, and realizes the intelligent translation of Chinese text information; for professional terms and vocabulary in specific fields, accurate translation is carried out through a customized domain term library.

[0050] Figure 1 In the technical solution provided by the illustrated embodiment, in step S120: According to the large semantic understanding model, combining a preset multilingual corpus and a domain term library, performing intelligent translation and domain term adaptation on Chinese text information to obtain internationalized text information, further includes: S130: According to the cultural environment applicable to the internationalized text information, performing cultural adaptation on the internationalized text information, multimedia resource information, and web page architecture information to obtain an internationalized resource file.

[0051] The technical solution provided by the embodiments of the present application can fully consider the cultural differences of different countries and regions, perform cultural adaptation processing on the translated content, such as replacing culturally sensitive vocabulary, adjusting the styles of pictures and icons, etc., to ensure that the web page content can be accurately understood and accepted in different cultural environments.

[0052] Specifically, as a preferred embodiment, as Figure 5 shown, in the above Chinese web page internationalization adaptation method, step S130: According to the cultural environment applicable to the internationalized text information, performing cultural adaptation on the internationalized text information, multimedia resource information, and web page architecture information to obtain an internationalized resource file, includes: S131: Obtaining the cultural environments of different countries and regions to form a cultural feature library, and using the cultural features in the cultural feature library to train a multimodal analysis model to construct a cultural semantic space; wherein, the multimodal deep analysis model is a two-stream heterogeneous deep network architecture; S132: Using the text processing stream of the two-stream heterogeneous deep network architecture, parsing the internationalized text information according to the cultural semantic space, and extracting the culturally sensitive text features in the internationalized text information; S133: Using the visual processing stream of the two-stream heterogeneous deep network architecture, using an object detection algorithm combined with a multi-scale attention mechanism, respectively detecting the feature associations in the multimedia resource information and the web page architecture information to obtain the culturally sensitive visual features corresponding to the multimedia resource information or the web page architecture information; S134: Replacing the culturally sensitive text features according to the cultural semantic space and adjusting the display of the culturally sensitive visual features.

[0053] The technical solution provided by the embodiment of the present application can perform multimodal analysis on the above-mentioned internationalized text information, multimedia resource information, and web page architecture information using a multimodal analysis model. Specifically, the multimodal analysis model is a dual-stream heterogeneous deep network architecture, which can process the text processing stream and the visual processing stream respectively. For the text processing stream, the internationalized text information is parsed according to the cultural semantic space to extract culturally sensitive text features. Specifically, the word vectors in the internationalized text information can be matched with the word vectors in the cultural semantic space for feature similarity. If the similarity is higher than or equal to a predetermined threshold, the keyword corresponding to the word vector is considered a culturally sensitive text feature. Then, using the visual processing stream, by adding a multi-scale attention mechanism to the target detection algorithm in this network architecture, the feature associations in the multimedia resource information and the web page architecture information can be monitored, such as colors, items, or people associated with specific cultures in pictures, so as to obtain culturally sensitive visual features. By replacing the culturally sensitive text features according to the cultural semantic space in the above manner and adjusting the display of the culturally sensitive visual features accordingly, the international cultural environment can be adapted. Specifically: if it is determined that there are culturally sensitive words, the culturally sensitive words in the internationalized text information are replaced with the cultural feature library; using the cultural feature library, it is judged whether there is a sensitive display style in the multimedia resource information; if it is determined that there is a sensitive display style, the display style corresponding to the multimedia resource information is adjusted using the cultural feature library; and, using the cultural feature library, it is judged whether the layout style corresponding to the web page architecture information adapts to the local cultural environment; if it is determined that it does not adapt to the local cultural environment, the layout style of the web page architecture information is adjusted using the cultural feature library.

[0054] In summary, the technical solution provided by the above embodiment of the present application fully considers the cultural differences in different countries and regions, and performs cultural adaptation processing on the translated content, such as replacing culturally sensitive vocabulary, adjusting the styles of pictures and icons, etc., to ensure that the web page content can be accurately understood and accepted in different cultural environments.

[0055] In addition, after performing mapping translation, professional term processing, and cultural adaptation on the internationalized text information, multimedia resource information, and web page architecture information corresponding to the above Chinese web page, format adaptive adjustment is also required to avoid format mismatch problems.

[0056] Specifically, as a preferred embodiment, as Figure 6 shown, after the step S130: performing cultural adaptation on the internationalized text information, multimedia resource information, and web page architecture information to obtain an internationalized resource file in the above Chinese web page internationalization adaptation method, the following steps are further included: S301: Render the web page using common rendering methods by simulating multiple target resolutions respectively; S302: Switching the display effect of each internationalized resource file on the web page in turn; S303: When there is a problem with the display effect of the internationalized resource file on the web page, adjust the web page architecture information.

[0057] The technical solution provided by the embodiment of the present application can simulate a large number of different target resolutions, use common rendering methods to virtually render web pages, switch each language resource file in turn, check the text occlusion, incomplete display and overlap problems, and give suggestions for web page layout adjustment, font adjustment and translation content adjustment. After obtaining the user's confirmation instruction, perform corresponding modification operations (such as content reduction, etc.), so as to ensure that the web page can present good visual effects and user experience on all target devices and browsers.

[0058] Figure 1 The technical solution provided by the illustrated embodiment further includes the following steps after obtaining the internationalized resource file: S140: When a web page needs to be run, an internationalized resource file is loaded according to the current language environment to obtain an internationalized web page corresponding to the current language environment.

[0059] Specifically, as a preferred embodiment, Figure 7 As shown, in the above-mentioned Chinese web page internationalization adaptation method, step S140: loading the internationalization resource file according to the current language environment, and obtaining the internationalized web page corresponding to the current language environment includes: S141: When a web page needs to be run, an HTML script corresponding to the current web page is loaded, wherein the HTML script includes a current language environment.

[0060] S142: Using HTML script to parse internationalized text information in the internationalized resource file.

[0061] S143: Loading the internationalized resource file according to the current language environment to obtain an internationalized web page in the current language environment.

[0062] The technical solution provided by the embodiment of the present application specifically loads the HTML script of the system during the running of the web page, automatically parses all key values wrapped by the multilingual tag __i18n__, loads the corresponding internationalized resource file according to the current language environment, and replaces the key value with the corresponding language resource text. Specifically, the external HTML script is used to run the internationalized resource file (for example, English en.json, Traditional Chinese zh_CHT.json), find the corresponding multilingual tag, and then find the key value corresponding to the multilingual tag, find the corresponding foreign text according to the key value, and parse the HTML language through the web page to obtain the display of the corresponding internationalized web page.

[0063] In summary, for the Chinese web page internationalization adaptation method provided by the above embodiments of the present application, first, a natural language model and a web page parsing algorithm are used to parse the Chinese web page, so as to obtain the Chinese text information, multimedia resource information, and web page architecture information of the Chinese web page; then, a semantic understanding large model is used to perform intelligent translation and domain term adaptation on the Chinese text information in combination with a preset multilingual corpus and a domain term library, so as to obtain internationalized text information. Because the semantic understanding large model is used to perform intelligent translation and domain term adaptation on the Chinese text information after the above natural language parsing in combination with the preset multilingual corpus and the domain term library, the semantics and context in the Chinese text information can be accurately understood, and the cultural differences of different countries and regions can be fully considered, and the professional terms in related fields can be accurately adapted, so as to reduce the problems of content disorder and format mismatch in the adaptation process. In addition, cultural adaptation is performed on the internationalized text information, multimedia resource information, and web page architecture information according to the cultural environment to which the internationalized text needs to be applied, so as to meet the international cultural needs. Finally, when the web page needs to be run, the internationalized resource file is loaded according to the current language environment, and the internationalized web page corresponding to the current language environment can be obtained. Through the above method, the problems in the prior art that the web page internationalization tools generally lack intelligence, are difficult to accurately understand the semantics and context of Chinese web pages, and are difficult to meet the internationalization requirements can be solved.

[0064] In addition, the following embodiments of the present application also provide a product embodiment of the Chinese web page internationalization adaptation based on a large model. The beneficial effects of the following product embodiments are the same as those of the Chinese web page internationalization adaptation method based on a large model provided by the above method embodiments, and other technical features in the product embodiments are the same as those disclosed in the above method embodiments, and will not be elaborated here.

[0065] See Figure 8 , Figure 8 which is a schematic structural diagram of a Chinese web page internationalization adaptation device provided by an embodiment of the present application. As Figure 8 shown, the Chinese web page internationalization adaptation device includes: A web page parsing module 110, configured to parse a Chinese web page using a natural language model and a web page parsing algorithm, and obtain the Chinese text information, multimedia resource information, and web page architecture information of the Chinese web page; An information adaptation module 120, configured to perform intelligent translation and domain term adaptation on the Chinese text information according to a semantic understanding large model in combination with a preset multilingual corpus and a domain term library, so as to obtain internationalized text information; A cultural adaptation module 130, configured to perform cultural adaptation on the internationalized text information, multimedia resource information, and web page architecture information according to the cultural environment to which the internationalized text information is applicable, so as to obtain an internationalized resource file; A file loading module 140, which is used to load an internationalized resource file according to the current language environment when a web page needs to be run, so as to obtain an internationalized web page corresponding to the current language environment.

[0066] In summary, the Chinese web page internationalization adaptation solution provided by the above embodiments of the present application can achieve the following technical effects compared with the prior art: (1) Significantly improved efficiency: Through automated steps such as in-depth parsing, text processing, and language resource file generation, the cumbersome processes of traditional manual translation and separately developing web page versions are abandoned, greatly shortening the time for Chinese web page internationalization adaptation, enabling rapid response to market demands, and meeting the timeliness requirements of users for web page internationalization.

[0067] (2) Greatly reduced costs: It reduces the large amount of human and material resources input required for manual translation and multi-version development, reduces development costs and maintenance costs, and alleviates the economic burden on enterprises in the process of web page internationalization.

[0068] (3) High adaptation accuracy: By leveraging intelligent capabilities in combination with multi-language corpora, semantic understanding models, and custom term libraries, it can accurately understand the semantics and context of Chinese texts to achieve accurate translation; at the same time, it fully considers cultural differences for adaptation processing, avoiding information transmission deviations caused by translation errors or cultural conflicts, and ensuring the accuracy and readability of web page content in different internationalization environments.

[0069] (4) Comprehensive format adaptation: By virtually rendering the web page at the target resolution and checking for problems, it gives comprehensive suggestions for layout, font, and translation content adjustments, effectively solving format problems such as text occlusion, incomplete display, and overlap, ensuring that the web page can present good visual effects and user experiences on various devices and browsers, and improving user satisfaction and the usability of the web page.

[0070] (5) Flexible and intelligent operation: When the web page is running, the system can automatically load the corresponding resource file according to the current language environment, realizing the rapid replacement of key values with corresponding language resource texts without manual intervention, enabling the web page to flexibly adapt to different language environments, and enhancing the versatility and practicality of the system.

[0071] In addition, it should be noted that the above examples are only for understanding the present application and do not constitute a limitation on the method of the present application. Based on this technical concept, more forms of simple transformations are within the protection scope of the present application.

[0072] In addition, see Figure 9, the electronic device provided by this application includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method for internationalization adaptation of Chinese web pages based on large models based on object detection in any of the above embodiments.

[0073] Reference is made below Figure 9 , which shows a schematic structural diagram of an electronic device suitable for implementing the embodiments of this application. The electronic device in the embodiments of this application can include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistant), PADs (Portable Application Description), PMPs (Portable Media Player), vehicle terminals (such as vehicle navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 9 The device shown is only an example and should not impose any limitations on the functions and usage scope of the embodiments of this application.

[0074] As Figure 9 shown, the electronic device can include a processing device 1001 (such as a central processing unit, a graphics processing unit, etc.), which can execute various appropriate actions and processes according to the program stored in the read-only memory (ROM: Read Only Memory) 1002 or the program loaded from the storage device 1003 into the random access memory (RAM: Random Access Memory) 1004. In the RAM 1004, various programs and data required for the operation of the electronic device are also stored. The processing device 1001, the ROM 1002, and the RAM 1004 are connected to each other through a bus 1005. The input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can enable the electronic device to communicate with other devices wirelessly or wirelessly to exchange data. Although the figure shows a model construction device with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be alternatively implemented or had.

[0075] In particular, according to the embodiments disclosed in the present application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the embodiments disclosed in the present application include a computer program product that includes a computer program carried on a computer-readable medium, and the computer program contains program codes for executing the methods shown in the flowcharts. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by a processing device 1001, the above-mentioned functions defined in the methods of the embodiments disclosed in the present application are executed.

[0076] It should be understood that the various parts disclosed in the present application can be implemented by hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in a suitable manner in any one or more embodiments or examples.

[0077] As described above, the above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed in the present application can easily think of changes or substitutions, and all should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0078] The present application provides a computer-readable storage medium having computer-readable program instructions (i.e., computer programs) stored thereon, and the computer-readable program instructions are used to execute the method for internationalization adaptation of Chinese web pages based on a large model in the above embodiments.

[0079] The computer-readable storage medium provided by the present application can, for example, be a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or components, or any combination of the above. More specific examples of the computer-readable storage medium can include, but are not limited to: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM) or flash memory, optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. In this embodiment, the computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, device, or component. The program code contained on the computer-readable storage medium can be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0080] The above computer-readable storage medium carries one or more programs. When the one or more programs are executed by a model construction device, computer program code for performing the operations of the present application can be written in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, executed as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, by connecting through an Internet service provider via the Internet).

[0081] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagram can represent a module, a segment of a program, or a part of code that contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks can occur in a different order than that marked in the accompanying drawings. For example, two consecutive blocks shown can actually be executed substantially in parallel, and they can sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, as well as combinations of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system that performs the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0082] The modules described in the embodiments of the present application can be implemented in software or in hardware. Among them, the name of the module does not constitute a limitation to the unit itself in some cases.

[0083] Each embodiment in this specification is described in a progressive manner. The same or similar parts among the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments.

[0084] The above are only the embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application can have various changes and modifications. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for internationalizing and adapting Chinese web pages based on large models, characterized in that, Including: Parsing a Chinese web page using a natural language model and a web page parsing algorithm to obtain the Chinese text information, multimedia resource information, and web page architecture information of the Chinese web page; According to a semantic understanding large model, combining a preset multilingual corpus and a domain term library, performing intelligent translation and domain term adaptation on the Chinese text information to obtain internationalized text information; According to the cultural environment applicable to the internationalized text information, performing cultural adaptation on the internationalized text information, the multimedia resource information, and the web page architecture information to obtain an internationalized resource file; When it is necessary to run the web page, loading the internationalized resource file according to the current language environment to obtain an internationalized web page corresponding to the current language environment.

2. The method according to claim 1, characterized in that, The step of parsing a Chinese web page using a natural language model and a web page parsing algorithm to obtain the Chinese text information, multimedia resource information, and web page architecture information of the Chinese web page includes: Parsing the Chinese web page using a natural language model to obtain the Chinese text information corresponding to the Chinese web page; Using the natural language model to split the Chinese text information into lexical and sentence forms, and analyzing the grammatical structure and semantic information of the lexical and sentence forms; Parsing the Chinese web page using a web page parsing algorithm to obtain the multimedia resource information and web page architecture information corresponding to the Chinese web page; Classifying and identifying the multimedia resource information.

3. The method according to claim 1, characterized in that, Before the step of performing intelligent translation and domain term adaptation on the Chinese text information according to a semantic understanding large model, combining a preset multilingual corpus and a domain term library to obtain internationalized text information, the method further includes: Refining the keywords in the Chinese text information as key values; Using the key values to replace the keywords in the Chinese text information and adding multilingual tags to the key values to obtain new Chinese text information; Generating a Chinese resource file containing the new Chinese text information; Inputting the Chinese resource file into the semantic understanding large model.

4. The method according to claim 1 or 2, characterized in that, The step of performing intelligent translation and domain term adaptation on the Chinese text information according to a semantic understanding large model, combining a preset multilingual corpus and a domain term library to obtain internationalized text information includes: Constructing a general semantic space using the multilingual corpus and training the semantic understanding large model using the multilingual corpus so that the semantic understanding large model learns the general semantic rules corresponding to the multilingual corpus; Inputting the Chinese text information into the semantic understanding large model to learn the grammatical structure and semantic information corresponding to the Chinese text information, where the Chinese text information includes keywords; Controlling the semantic understanding large model, combining the grammatical structure and semantic information, to learn the semantic association relationship between the keywords and the context and capture the semantic roles corresponding to the keywords; Controlling the semantic understanding large model, according to the semantic association relationship and the semantic roles, using the general semantic rules to map the keywords to general word vectors in the semantic space; Moreover, construct a professional semantic space using the domain term library, and train the semantic understanding large model using the domain term library so that the semantic understanding large model learns the professional semantic rules corresponding to the domain term library; Control the semantic understanding large model to map keywords containing professional domain terms to professional word vectors in the professional semantic rules according to the professional semantic rules; According to the mapping relationship between Chinese and other language texts in the multilingual corpus, map the general word vectors and professional word vectors to foreign language texts respectively to obtain the internationalized text information.

5. The method according to claim 1 or 2, characterized in that The step of performing cultural adaptation on the internationalized text information, the multimedia resource information, and the web page architecture information according to the cultural environment applicable to the internationalized text information to obtain an internationalized resource file includes: Obtain the cultural environments of different countries and regions to form a cultural feature library, and use the cultural features in the cultural feature library to train a multimodal analysis model to construct a cultural semantic space; wherein, the multimodal depth analysis model is a dual-stream heterogeneous deep network architecture; Use the text processing stream of the dual-stream heterogeneous deep network architecture to parse the internationalized text information according to the cultural semantic space, and extract the culturally sensitive text features in the internationalized text information; Use the visual processing stream of the dual-stream heterogeneous deep network architecture, and use an object detection algorithm combined with a multi-scale attention mechanism to detect the feature associations in the multimedia resource information and the web page architecture information respectively to obtain the culturally sensitive visual features corresponding to the multimedia resource information or the web page architecture information; Replace the culturally sensitive text features according to the cultural semantic space, and adjust the display of the culturally sensitive visual features.

6. The method according to claim 1, characterized in that After the step of performing cultural adaptation on the internationalized text information, the multimedia resource information, and the web page architecture information to obtain an internationalized resource file, the method further includes: Simulate rendering the web page using common rendering methods at multiple target resolutions respectively; Switch the display effects of each internationalized resource file on the web page in sequence; When there is a problem with the display effect of the internationalized resource file on the web page, adjust the web page architecture information.

7. The method according to claim 1, wherein The step of loading the internationalized resource file according to the current language environment to obtain an internationalized web page corresponding to the current language environment includes: When the web page needs to be run, load the HTML script corresponding to the current web page, wherein the HTML script includes the current language environment; Use the HTML script to parse the internationalized text information in the internationalized resource file; Load the internationalized resource file according to the current language environment to obtain an internationalized web page in the current language environment.

8. An internationalization adaptation device for Chinese web pages based on a large model, characterized in that including: A web page parsing module for parsing a Chinese web page using a natural language model and a web page parsing algorithm to obtain the Chinese text information, multimedia resource information, and web page architecture information of the Chinese web page; An information adaptation module for performing intelligent translation and domain term adaptation on the Chinese text information according to a semantic understanding large model in combination with a preset multilingual corpus and a domain term library to obtain internationalized text information; A cultural adaptation module, configured to perform cultural adaptation on the internationalized text information, the multimedia resource information, and the web page architecture information according to the cultural environment applicable to the internationalized text information, so as to obtain an internationalized resource file; A file loading module, configured to load the internationalized resource file according to the current language environment when the web page needs to be run, so as to obtain an internationalized web page corresponding to the current language environment.

9. A computer storage medium having computer-executable instructions stored thereon, characterized in that, When the computer program is executed by a processor, it implements the large model-based Chinese web page internationalization adaptation method according to any one of claims 1 to 7.

Images