WO2021227951A1

WO2021227951A1 - Naming of front-end page element

Info

Publication number: WO2021227951A1
Application number: PCT/CN2021/092136
Authority: WO
Inventors: 谢杨易; 崔恒斌
Original assignee: 支付宝(杭州)信息技术有限公司
Priority date: 2020-05-09
Filing date: 2021-05-07
Publication date: 2021-11-18
Also published as: CN112307235B; CN111291208A; CN112307235A; CN111291208B

Abstract

A front-end page element naming method and apparatus, and an electronic device. The method comprises: if a target page element is an image element, calculating the similarity between the target page element and each image in a preset image database (S102); determining the maximum similarity in the calculated similarities (S104); and determining the name of the image in the preset image database corresponding to the calculated maximum similarity as the name of the target page element (S106).

Description

Naming of front-end page elements

Technical field

This application relates to the field of computer network technology, and in particular to a method, device and electronic equipment for naming front-end page elements.

Background technique

In the front-end page development work, in order to help improve the readability of the front-end page code and the convenience of maintaining the code in the later stage, developers usually need to name the front-end page elements.

Currently, when naming page elements, developers usually need to manually name them. However, because element naming has strict specifications, and the number of page elements included in front-end development is large, the manual naming of elements may lead to low naming efficiency, failure to strictly abide by naming conventions, and naming errors.

Summary of the invention

This application proposes a method for naming front-end page elements. The above method includes: when the target page element is an image element, calculating the similarity between the target page element and each image in the preset image library; determining the calculated The maximum similarity in the above-mentioned similarity; the name of the image in the above-mentioned preset image library corresponding to the calculation of the above-mentioned maximum similarity is determined as the name of the target page element.

In the illustrated embodiment, calculating the similarity between the target page element and each image in the preset image library includes: inputting the element data of the target page element into a pre-trained classification model for calculation , Obtain the classification result of the target page element, the classification model is a neural network model trained based on a number of samples labeled with the classification result; search for an image that is the same as the classification result of the target page element from the preset image library; Calculate the similarity between the above-mentioned target page elements and the searched images.

In the illustrated embodiment, the above method further includes: when the target page element is a text element, inputting the element data of the target page element into a pre-trained translation model for calculation to obtain the English language corresponding to the target page element String; the above-mentioned English string is determined as the name of the above-mentioned target page element.

In the illustrated embodiment, the above method further includes: converting the traditional characters in the target page elements into simplified characters based on a pre-built mapping algorithm.

In the illustrated embodiment, determining the above-mentioned English character string as the name of the target page element includes: inputting the above-mentioned English character string into a pre-trained keyword extraction model for calculation, and obtaining the same as the above-mentioned English character string. Corresponding keywords; the above keywords are determined as the names of the above target page elements.

In the illustrated embodiment, the above method further includes: if the target page element is a container element, adding an identifier indicating that the target page element is a container element to the name of the target page element.

In the illustrated embodiment, adding an identifier indicating that the target page element is a container element to the name of the target page element includes: extracting keywords from the names of each element in the container element; Words are combined to obtain the name of the target page element; an identifier indicating that the target page element is a container element is added to the name.

This application also proposes a device for naming front-end page elements, including: a calculation module, when the target page element is an image element, calculate the similarity between the target page element and each image in the preset image library; first The determining module determines the maximum similarity among the calculated similarities; the second determining module determines the name of the image in the preset image library corresponding to the calculation of the maximum similarity as the name of the target page element.

In the illustrated embodiment, the calculation module includes: inputting the element data of the target page element into a pre-trained classification model for calculation to obtain the classification result of the target page element, and the classification model is based on a number of labeled The neural network model obtained by training the sample of the classification result; searching for the image that is the same as the classification result of the target page element from the preset image library; calculating the similarity between the target page element and the searched images.

In the illustrated embodiment, the above-mentioned device further includes: a model calculation module, when the target page element is a text element, the element data of the above-mentioned target page element is input into a pre-trained translation model for calculation, and the result is the same as that of the above-mentioned target page. The English character string corresponding to the element; the third determining module determines the above English character string as the name of the target page element.

In the illustrated embodiment, the above device further includes: a conversion module, which converts the traditional characters in the target page elements into simplified characters based on a pre-built mapping algorithm.

In the illustrated embodiment, the third determining module includes: inputting the English character string into a pre-trained keyword extraction model for calculation to obtain keywords corresponding to the English character string; Determine the name of the above target page element.

In the illustrated embodiment, the above-mentioned apparatus further includes: an adding module, if the above-mentioned target page element is a container element, add an identifier indicating that the above-mentioned target page element is a container element to the name of the above-mentioned target page element.

In the illustrated embodiment, the above-mentioned adding module includes: extracting keywords from the names of each element in the above-mentioned container element; combining the various keywords to obtain the name of the above-mentioned target page element; adding to the above-mentioned name The identifier indicating that the above target page element is a container element.

It can be seen from the above technical solution that, on the one hand, when the element is an image element, the above system can calculate the similarity between the target page element and each image in the preset image library, and combine the preset image library with The name of the image corresponding to the maximum similarity in the calculated similarity is determined as the name of the target page element.

On the other hand, when the element is a text element, the above-mentioned system can extract keywords from the above-mentioned text element, and use the extracted keywords as the name of the above-mentioned text element.

On the other hand, when the element is a container element, the above-mentioned system may add an identifier indicating that the above-mentioned target page element is a container element to the name of the above-mentioned target page element, so as to realize the naming of the above-mentioned container element.

Therefore, the element naming method disclosed in this application can realize automatic naming of elements, thereby improving element naming efficiency, element naming standardization, and correctness, avoiding low naming efficiency due to manual participation and failing to strictly comply with naming conventions when naming. Problems such as naming errors.

Description of the drawings

FIG. 1 is a method flowchart of a method for naming front-end page elements shown in this application;

FIG. 2 is a method flowchart of the text element naming method shown in this application;

FIG. 3 is a method flowchart of the container element naming method shown in this application;

FIG. 4 is a structural diagram of a device for naming front-end page elements shown in this application;

Fig. 5 is a hardware structure diagram of a naming device for front-end page elements shown in this application.

Detailed ways

The exemplary embodiments will be described in detail below, and examples thereof are shown in the accompanying drawings. When the following description refers to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all implementation manners consistent with the present application. Rather, they are merely examples of devices and methods consistent with some aspects of the application as detailed in the appended claims.

The terms used in this application are only for the purpose of describing specific embodiments, and are not intended to limit the application. The singular forms of "a", "above" and "the" used in this application and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and/or" as used herein refers to and includes any or all possible combinations of one or more associated listed items. It should also be understood that the word "if" used herein, depending on the context, can be interpreted as "when" or "when" or "in response to determination".

This application aims to propose a method for naming front-end page elements, so that when determining the name of the page element, the page element name determination system realizes the naming of different types of page elements, thereby avoiding low naming efficiency due to manual participation , The naming cannot strictly abide by the naming convention, naming errors and other issues.

The technical solutions disclosed in the present application will be described below in conjunction with specific embodiments.

Please refer to FIG. 1. FIG. 1 is a method flowchart of a method for naming front-end page elements shown in this application. Applied to the page element naming system. As shown in Figure 1, the above method includes:

S102: When the target page element is an image element, calculate the similarity between the target page element and each image in the preset image library.

S104: Determine the maximum similarity among the calculated similarities.

S106: Determine the name of the image in the preset image library corresponding to the calculation of the maximum similarity as the name of the target page element.

The above-mentioned page element naming system (hereinafter referred to as the "system") may specifically be a piece of logic code carried in a terminal device. The above-mentioned page element naming system needs to provide computing power through its equipped terminal device when executing the above-mentioned element extraction method as the execution subject.

In practical applications, the above system can provide an interactive platform for interacting with developers. Through this interactive platform, on the one hand, developers can provide the page elements that need to be named to the above system, and initiate related instructions for naming the page elements to the above system; on the other hand, when the page elements are named, the above The system can output the named page elements to developers.

The above-mentioned front-end page image is specifically a page image designed by a page image designer. In actual situations, when a developer develops a front-end page, he usually needs to refer to the page image designed by the page image designer for development, so that the display effect of the final developed front-end page can be the same as the above-mentioned page image.

The above-mentioned front-end page elements (hereinafter referred to as "elements") specifically constitute the main components of the front-end page, which may include image elements, text elements, and container elements.

The above-mentioned image element specifically refers to an element whose content is an image.

The above text element specifically refers to an element whose content is text. Among them, the above-mentioned characters may include traditional or simplified characters.

The aforementioned container element specifically refers to a collection of elements composed of several elements. In practical applications, several image elements can form a container element. Several text elements can form a container element. Several text elements and several image elements can also form a container element together.

It is understandable that in practical applications, the naming conventions for different types of elements are different. Therefore, when naming elements, you need to determine the element type of the element.

In one embodiment, when a developer needs to name a certain element, the developer can provide the above-mentioned element and the element type of the element to the above-mentioned system through the interactive platform provided by the above-mentioned system.

For example, the aforementioned interactive platform may provide a window for the developer to input the element type of the element to be named. When the developer provides the element data of the aforementioned element to the aforementioned system, the developer can also input the element type of the aforementioned element in the aforementioned window for the aforementioned system to identify the element type.

In another embodiment, in order to improve the efficiency and accuracy of element naming. When a developer needs to name an element, the developer can only provide the aforementioned element to the aforementioned system through the interactive platform provided by the aforementioned system.

In the above-mentioned situation, the above-mentioned system can automatically identify the element type of the above-mentioned element.

In one implementation, when identifying the element type of the above-mentioned element, the above-mentioned system may first perform OCR recognition on the element data corresponding to the above-mentioned element to obtain the recognition result corresponding to the above-mentioned element, and then determine each element according to the above-mentioned recognition result. The element type.

Before introducing the specific steps, this application first introduces the principle of determining element types through OCR identification.

OCR (Optical Character Recognition, Optical Character Recognition) technology is specifically a technology that directly converts text content on images and photos into editable text. The principle is to compare the image features of the target image with the image features of Chinese characters in the existing Chinese character library, and output the Chinese character that most matches the image features of the target image as the recognition result, and the recognition confidence of the above recognition result. Wherein, the recognition confidence level may indicate to a certain extent the similarity between the image feature of the target image and the recognition result.

For example, if the text content included in the target image is the Chinese character "中", at this time, since the target image does include a Chinese character, the recognition confidence of the recognition result obtained after the OCR detection will be relatively high. Suppose that the specific content of the target image is a pattern similar to the Chinese character "中". At this time, after OCR recognition is performed on the target image, although the corresponding recognition result can be obtained, the specific content included in the above target image is only similar to Chinese characters. Therefore, the above-mentioned recognition confidence will be relatively low.

It can be seen that when the element type of the above element is determined by OCR recognition, after OCR recognition is performed on the element image of the element, it is possible to determine whether the recognition confidence corresponding to the recognition result reaches a preset threshold. Element type. The above-mentioned preset threshold may be specifically set by the developer based on experience or trained through a large number of samples, which is not limited here. When the recognition confidence level reaches the preset threshold, the element type of the element is determined to be a text element; otherwise, the element type of the element is determined to be an image element.

It is understandable that, in the above-mentioned situation, if several recognition results are obtained after OCR recognition for the above-mentioned element, it means that the above-mentioned element is a collection of several text elements or image elements. At this time, it can be determined that the above-mentioned element is a container element.

In another embodiment, when determining the element type of the element, the above system may input the element data corresponding to the element into a pre-trained classifier for calculation, and determine the element type of the element based on the calculation result.

Among them, the above-mentioned classifier may be specifically obtained by training based on a number of element image samples marked with element types; the above-mentioned element types include image elements, text elements, and container elements.

It should be noted here that the structure and type of the above-mentioned classifier are not limited here. The above-mentioned classifier may be a multi-classifier constructed based on a neural network.

The aforementioned image library may specifically be a pre-configured image library. The aforementioned image library can usually include several named images (images named according to naming conventions).

In practical applications, in order to standardize the storage of images and improve the accuracy of naming, the images included in the above-mentioned image library can be classified and stored. For example, the above-mentioned image library can be divided into several storage spaces; among them, each storage space can store images of the same image type.

In a way of configuring the image library, developers can obtain an image collection that includes several common element images. Then, the developer can name each image in the above-mentioned image collection according to the naming convention, classify the named images (either manually or through a classifier), and save them in the storage space corresponding to the above-mentioned image library. It is understandable that the configured image library can be duplicated and used repeatedly, and it does not need to be configured every time the target element is named. Of course, the configured image library can be updated. For example, adding a new image or updating the name of an existing image, etc.

When it is determined that the target element is an image element, the foregoing system may execute S102 to calculate the similarity between the foregoing target element and each image in the preset image library.

In an embodiment, when calculating the similarity between the target element and each image in the preset image library, the system may first organize the element data of the target element into the form of a feature vector, so as to facilitate similarity.度Calculation.

For example, the above-mentioned system may first extract the image features of the above-mentioned target elements (for example, Harris corner points or SIFT features), and form corresponding feature vectors.

After that, the foregoing system may execute the following steps S1022-S1026 for each image in the foregoing preset image library:

S1022: Extract image features of the aforementioned image to form a feature vector.

S1024: After extracting the feature vector, calculate the Euclidean distance between each feature vector corresponding to the image and each feature vector corresponding to the target element, and count the number of feature vectors whose Euclidean distance is less than a preset reference threshold.

S1026: In the statistical feature vectors included in the image, the Euclidean distance between the feature vector included in the target element and the feature vector is less than a preset reference threshold, and the preset mapping algorithm (for example, normalization or standardization algorithm) is used. , The above-mentioned quantity is mapped to the similarity between the above-mentioned image and the above-mentioned target element.

S1028: Record the above-mentioned similarity of the mapping and the corresponding relationship with the above-mentioned image.

Here, it should be noted that the method for calculating the similarity is not limited in this application. For example, the above method for calculating similarity can also be calculated by calculating the cosine distance, Manhattan distance, Mahalanobis distance between feature vectors.

After completing the above steps for each image in the preset image library, the system will obtain the similarity between the target element and each image, as well as the similarity, and the corresponding relationship with each image.

Then, the system can execute S104-S106 to determine the maximum similarity among the calculated similarities, and determine the name of the image in the preset image library corresponding to the calculation of the maximum similarity as the target page element The name.

In one embodiment, in order to improve the efficiency of determining the above-mentioned maximum similarity, the above-mentioned system can push the obtained above-mentioned similarity into the big top pile (the value corresponding to each parent node in the big top pile is greater than or equal to its left and right children). The value corresponding to the node). Then, the system can read the similarity stored in the root node of the big top pile, and determine the read similarity as the maximum similarity.

It is not difficult to understand that because the characteristic of the big top heap is that the value corresponding to each parent node is greater than or equal to the value corresponding to its left and right child nodes, therefore, the root node of the big top heap records the maximum value maintained in the above big top heap . It can be seen that the similarity stored in the root node of the above-mentioned large top pile is the maximum similarity among the obtained similarities.

After determining the maximum similarity, the system can determine the image corresponding to the maximum similarity from the recorded correspondence. After determining the image, the system may determine the name of the image as the name of the target element.

At this point, the above system has completed the naming of the target element.

It can be seen from the above technical solution that, when naming front-end page elements, the above system can calculate the similarity between the target page element and each image in the preset image library, and combine the preset image library with The name of the image corresponding to the maximum similarity in the calculated similarity is determined as the name of the target page element mentioned above. Therefore, it is possible to automatically name the element, thereby improving the efficiency of element naming, the standardization of element naming, and the correctness, avoiding The naming efficiency is low due to manual participation, the naming cannot be strictly abided by the naming convention, naming errors and other issues.

In one embodiment, in order to improve the naming accuracy, when the system executes S102 to calculate the similarity between the target page element and each image in the preset image library, the element data of the target page element may be first Input the pre-trained classification model for calculation, and obtain the classification result of the target page element.

Among them, the above classification model is a neural network model obtained by training based on a number of samples labeled with classification results.

When training the above classification model, a number of sample data marked with classification results can be obtained first. After obtaining several sample data, the above sample data can be input into the classification model, and iterative training can be performed until the above classification model converges. At this time, the convergent classification model can be used as the trained classification model.

After determining the image type of the target element, the system can search for images that are the same as the classification result of the target page element from the preset image library, and then calculate the target page element to be similar to the searched images Spend.

In one manner, when searching for an image that is the same as the classification result of the target page element, the system may directly read the image recorded in the storage space corresponding to the classification result.

In another way, when searching for images that are the same as the classification result of the target page element, the system may input the image data of each image in the preset image library into the classification model for calculation, thereby obtaining the image data of each image. Image type. After that, the above-mentioned system may determine an image whose image type is the same as the image type of the above-mentioned target element as an image with the same classification result of the above-mentioned target page element.

After determining the similarity between the target element and each image, the system can continue to execute S104-S106 to determine the maximum similarity among the calculated similarities; the preset image corresponding to the maximum similarity will be calculated The name of the image in the library is determined as the name of the above-mentioned target page element (for detailed steps, please refer to the foregoing content, which will not be described in detail here).

In this embodiment, because the system described above determines the image that is most similar to the target element among the images of the same type as the target element, and uses the name of the most similar image as the name of the target element, therefore, Can improve the accuracy of element naming.

Please refer to FIG. 2. FIG. 2 is a method flowchart of the text element naming method shown in this application.

When it is determined that the target element is a text element, as shown in FIG. 2, the above-mentioned system can first convert the text content of the above-mentioned text element into traditional and simplified form.

In practical applications, the above-mentioned system can be equipped with a mapping algorithm for converting traditional characters to simplified characters in advance. Through this mapping algorithm, the above-mentioned system can convert traditional characters in text elements into simplified characters.

For example, the above-mentioned mapping algorithm may be an algorithm for converting traditional characters to simplified characters constructed based on the hanlp tool. After receiving the text content of the text element, the algorithm can first segment the above text content according to the text, and then detect whether the divided text is a traditional Chinese character one by one, and if it is, it will be converted to the corresponding simplified character for output; if Otherwise, output the divided text directly. After the traditional and simplified conversion is performed for each divided group, the above algorithm can recombine the output simplified characters into the text content of the above text element.

Here, it should be noted that this application does not limit the foregoing mapping algorithm.

After obtaining the text element that has been converted from traditional to simplified, the system can input the element data of the target page element into a pre-trained translation model for calculation to obtain the English character string corresponding to the target page element.

In practical applications, the above-mentioned system can be pre-loaded with a trained translation model. Through this translation model, the input Chinese text content can be converted into English text content.

For example, the aforementioned translation model may be an NLP (Natural Language Processing) model based on seq2seq. After receiving the text content of the text element, the model can first segment the text content according to the text, and then use the segmented text as input for semantic encoding to obtain the vector corresponding to the text content. After the semantic coding is completed, the above-mentioned vector can be decoded into English text content based on the above-mentioned semantic coding and the English word library.

Here, it should be noted that this application does not limit the above-mentioned translation model.

After the text content of the text element is converted into English text content (the text content is composed of English strings), the system can select several keywords from the English text content as the name of the text element.

In practical applications, the above-mentioned system can be equipped with a keyword extraction model in advance. Through the keyword extraction model, keywords can be extracted from the input English text content.

For example, the above keyword extraction model may be a model constructed based on the TF-IDF algorithm. After receiving the English text content of the text element, the model can first segment each word in the above-mentioned English text, and then count the occurrence frequency (TF, Term Frequency, word frequency) of the segmented word in the text. After calculating the frequency of each word in this text, you can combine the frequency of each word in other English texts (IDF, Inverse Documnet Frequency, inverse document frequency), sort the words, and rank them in the top N Bit words are used as keywords; among them, N is a positive integer preset based on experience.

For another example, the keyword extraction model may be an NLP model based on textRank. After receiving the English text content of the text element, the model can first segment each word in the above-mentioned English text. After obtaining the divided words, the above system can combine two adjacent divided words in pairs to obtain all possible combinations, and then calculate the connection weights between the words in the combination. After calculating the connection weights between words in each combination, the above-mentioned system can calculate the sum of the connection weights corresponding to each word, and sort the words in the above-mentioned English text according to the size of the above-mentioned sum. At this time, the above system can use the top N words as keywords; where N is a positive integer preset based on experience.

Here, it should be noted that this application does not limit the above-mentioned keyword extraction model.

After the keyword is determined from the English text content of the text element, the system may determine the keyword as the name of the text element.

When it is determined that the target element is a container element, the foregoing system may add an identifier indicating that the target page element is a container element to the name of the target page element.

Please refer to FIG. 3, which is a method flowchart of the container element naming method shown in this application.

When it is determined that the target element is a container element, as shown in Figure 3, the above-mentioned system can first determine the element type of each element included in the above-mentioned container element.

In practical applications, the above system can use the method for determining element types disclosed in this application to determine the element types of the above elements one by one.

When the above-mentioned container element only includes the only text element, the above-mentioned system can use the naming method for text elements disclosed in this application to name the text element in the above-mentioned container element. After the naming is completed, the system may add an identifier indicating that the target page element is a container element to the name of the text element as the name of the container element. For example, add the characters "contarner" before the name of the above text element.

When the container element includes several text elements, in one embodiment, the system may first determine the text element for naming from the container element. Then, the above-mentioned system can use the naming method for text elements disclosed in this application to name the determined text elements. After the naming is completed, the system may add an identifier indicating that the target page element is a container element to the name of the text element as the name of the container element.

For example, the aforementioned system may determine the first (last) text element in the aforementioned container element as the text element for naming, and perform subsequent naming.

For another example, the foregoing system may determine the text element with the largest amount of data among the foregoing container elements as the text element for naming, and perform subsequent naming.

For another example, the above text element carries an identifier indicating the importance of the above text element (the larger the value indicated by the identifier, the higher the importance of the above text element). The foregoing system may determine the text element with the largest value of the foregoing identifier carried in the foregoing container element as the text element for naming, and perform subsequent naming.

Here, it should be noted that the method for determining the text element used for naming can be set according to the actual situation, which is not limited here.

In another embodiment, the above-mentioned system may first use the method for naming text elements disclosed in this application to extract the keywords of each text element. Then, the above-mentioned system may combine the keywords to obtain the combined keywords, and add an identifier indicating that the target page element is a container element to the combined keywords as the name of the container element.

In another embodiment, the above-mentioned system may first use the method for naming text elements disclosed in this application to extract the keywords of each text element. Then, the system can determine the most important keyword among the proposed keywords, and add an identifier indicating that the target page element is a container element to the most important keyword as the name of the container element.

For example, when determining the above-mentioned most important keywords, the above-mentioned system may input each keyword into the keyword extraction model described in this application for calculation, and then use the calculation result as the above-mentioned most important keyword.

When the aforementioned container element only includes a unique image element, the aforementioned system can use the naming method for image elements disclosed in this application to name the text element in the aforementioned container element. After the naming is completed, the system may add an identifier indicating that the target page element is a container element to the name of the image element as the name of the container element. For example, add the characters "contarner" before the name of the above text element.

When the container element includes several image elements, in one embodiment, the system may first determine the image element used for naming from the container elements. Then, the above system can use the naming method for text elements disclosed in this application to name the determined image elements. After the naming is completed, the system may add an identifier indicating that the target page element is a container element to the name of the image element as the name of the container element.

For example, the above-mentioned system may determine the first (last) image element among the above-mentioned container elements as the image element for naming, and perform subsequent naming.

For another example, the foregoing system may determine the image element with the largest amount of data among the foregoing container elements as the image element for naming, and perform subsequent naming.

For another example, the above-mentioned image element carries an identifier indicating the importance of the above-mentioned image element (the larger the value indicated by the identifier, the higher the importance of the above-mentioned image element). The above-mentioned system may determine the image element with the largest value of the above-mentioned identifier carried in the above-mentioned container element as the image element for naming, and perform subsequent naming.

Here, it should be noted that the method for determining the image element used for naming can be set according to the actual situation, which is not limited here.

In another embodiment, the aforementioned system may first use the method for naming image elements disclosed in this application to determine the name of each image element. Then, the above system may combine the names of the image elements to obtain a combined name, and add an identifier indicating that the target page element is a container element to the combined name as the name of the container element.

In another embodiment, the aforementioned system may first use the method for naming image elements disclosed in this application to determine the name of each image element. Then, the above-mentioned system may extract keywords from the determined names of each image element, and add an identifier indicating that the target page element is a container element to the above-mentioned keywords as the name of the container element.

For example, when extracting the above-mentioned keywords, the above-mentioned system may input the name of each image element into the keyword extraction model described in this application for calculation, and then use the calculation result as the above-mentioned keyword.

It should be noted that when the aforementioned container element includes both text elements and image elements, the naming method of the aforementioned container element can refer to the aforementioned content, which will not be described in detail here.

When the container element does not include any element, the system may combine the identifier indicating that the target page element is the container element with the assigned sequence number of the container element, and use the combined result as the name of the container element.

It should be noted that the sequence numbers assigned to the above-mentioned container elements may be assigned according to actual conditions, which are not limited here. For example, in one situation, the sequence number to which the above-mentioned container elements are assigned may indicate the order in which the above-mentioned container elements are created. In another case, the sequence number assigned to the above-mentioned container element may be a sequence number assigned manually.

Correspondingly, this application also proposes a device for naming front-end page elements. Please refer to FIG. 4, which is a structural diagram of a device for naming front-end page elements shown in this application.

As shown in FIG. 4, the foregoing apparatus 400 may include:

The calculation module 410, when the target page element is an image element, calculates the similarity between the target page element and each image in the preset image library;

The first determining module 420 determines the maximum similarity among the calculated similarities;

The second determining module 430 determines the name of the image in the preset image library corresponding to the calculation of the maximum similarity as the name of the target page element.

In the illustrated embodiment, the calculation module 410 includes: inputting the element data of the target page element into a pre-trained classification model for calculation to obtain the classification result of the target page element; wherein, the classification model is based on A neural network model trained by a number of samples labeled with classification results; from a preset image library, search for images that are the same as the classification results of the target page elements; calculate the difference between the target page elements and the searched images Similarity.

In the illustrated embodiment, the above-mentioned device 400 further includes: a model calculation module, when the target page element is a text element, the element data of the above-mentioned target page element is input into a pre-trained translation model for calculation, and the result is the same as the above-mentioned target page element. The English character string corresponding to the page element; the third determining module determines the above-mentioned English character string as the name of the above-mentioned target page element.

In the illustrated embodiment, the device 400 further includes a conversion module, which converts the traditional characters in the target page elements into simplified characters based on a pre-built mapping algorithm.

In the illustrated embodiment, the apparatus 400 further includes: an adding module, if the target page element is a container element, add an identifier indicating that the target page element is a container element to the name of the target page element.

The embodiment of the device for naming front-end page elements shown in this application can be applied to a device for naming front-end page elements. The device embodiments can be implemented by software, or can be implemented by hardware or a combination of software and hardware. Taking software implementation as an example, as a logical device, it is formed by reading the corresponding computer program instructions in the non-volatile memory into the memory through the processor of the electronic device where it is located. From a hardware perspective, as shown in Figure 5, the hardware structure diagram of a naming device for front-end page elements shown in this application, except for the processor, memory, network interface, and non-volatile memory shown in Figure 5 In addition to the memory, the electronic device in which the device is located in the embodiment usually includes other hardware according to the actual function of the electronic device, which will not be repeated here.

Please refer to a device for naming front-end page elements shown in FIG. 5. The device includes: a processor; and a memory for storing executable instructions of the processor. The above-mentioned processor is configured to call the executable instructions stored in the above-mentioned memory to implement any one of the above-mentioned methods for naming front-end page elements.

It should be understood that the present application is not limited to the precise structure that has been described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is only limited by the attached authority requirements.

The above are only preferred embodiments of this application, and are not intended to limit this application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this application shall be included in the protection of this application. Within the range.

Claims

The naming methods of front-end page elements include:

When the target page element is an image element, calculating the similarity between the target page element and each image in the preset image library;

Determining the maximum similarity among the calculated similarities;

The name of the image in the preset image library corresponding to the calculation of the maximum similarity is determined as the name of the target page element.
The method according to claim 1, wherein the calculating the similarity between the target page element and each image in a preset image library comprises:

Input the element data of the target page element into a pre-trained classification model for calculation to obtain the classification result of the target page element; wherein, the classification model is a neural network trained based on a number of samples labeled with classification results Model;

Search for an image that is the same as the classification result of the target page element from the preset image library;

Calculate the similarity between the target page element and the found images.
The method according to claim 1, further comprising:

When the target page element is a text element, input the element data of the target page element into a pre-trained translation model for calculation to obtain an English character string corresponding to the target page element;

The English character string is determined as the name of the target page element.
The method according to claim 3, further comprising:

Based on a pre-built mapping algorithm, the traditional characters in the target page elements are converted into simplified characters.
The method according to claim 3, wherein the determining the English character string as the name of the target page element includes:

Input the English character string into a pre-trained keyword extraction model for calculation to obtain keywords corresponding to the English character string;

The keyword is determined as the name of the target page element.
The method according to any one of claims 1-5, further comprising:

If the target page element is a container element, an identifier indicating that the target page element is a container element is added to the name of the target page element.
The method according to claim 6, wherein the adding an identifier indicating that the target page element is a container element to the name of the target page element comprises:

Extract keywords from the names of each element in the container element;

Combine the keywords to obtain the name of the target page element;

An identifier indicating that the target page element is a container element is added to the name.
The naming device of front-end page elements includes:

A calculation module, when the target page element is an image element, calculate the similarity between the target page element and each image in the preset image library;

The first determining module determines the maximum similarity among the calculated similarities;

The second determining module determines the name of the image in the preset image library corresponding to the calculation of the maximum similarity as the name of the target page element.
The device according to claim 8, the calculation module comprising:

Input the element data of the target page element into a pre-trained classification model for calculation to obtain the classification result of the target page element; wherein, the classification model is a neural network trained based on a number of samples labeled with classification results Model;

Search for an image that is the same as the classification result of the target page element from the preset image library;

Calculate the similarity between the target page element and the found images.
The device according to claim 8, further comprising:

A model calculation module, when the target page element is a text element, input the element data of the target page element into a pre-trained translation model for calculation to obtain an English character string corresponding to the target page element;

The third determining module determines the English character string as the name of the target page element.
The device according to claim 10, further comprising:

The conversion module converts the traditional characters in the target page elements into simplified characters based on a pre-built mapping algorithm.
The device according to claim 10, the third determining module comprises:

Input the English character string into a pre-trained keyword extraction model for calculation to obtain keywords corresponding to the English character string;

The keyword is determined as the name of the target page element.
The device according to any one of claims 8-12, further comprising:

An adding module, if the target page element is a container element, add an identifier indicating that the target page element is a container element to the name of the target page element.
The device according to claim 13, wherein the adding module comprises:

Extract keywords from the names of each element in the container element;

Combine the keywords to obtain the name of the target page element;

An identifier indicating that the target page element is a container element is added to the name.
The naming device for front-end page elements, including:

processor;

A memory for storing executable instructions of the processor;

Wherein, the processor is configured to call executable instructions stored in the memory to implement the method for naming front-end page elements according to any one of claims 1 to 14.