Development component recommendation method and device, equipment and medium

Abstract

The invention discloses a development component recommendation method and device, equipment and a medium, and relates to the technical field of computer processing, and the method comprises the steps: determining design image data and project text information corresponding to a target development project, respectively analyzing the design image data and the project text information by utilizing a preset image processing model and a preset text processing model; integrating the obtained project visual features, the first text features and the second text features, and performing enhancement processing on the integrated features by using a preset feature enhancement algorithm; respectively determining a plurality of component characteristics corresponding to the plurality of development components, and performing evaluation; and sorting the plurality of development components according to the obtained plurality of target recommendation scores, and determining a to-be-recommended development component from the plurality of development components so as to carry out development component recommendation. Therefore, the problem of low development efficiency caused by insufficient utilization of development component resources and limitation of development architecture design can be solved while the product quality is ensured.

Description

Technical Field

[0001] The present invention relates to the field of computer processing technology, and in particular to a development component recommendation method, device, equipment and medium. Background Art

[0002] With the rapid development of Internet technology, front-end component development has become a mainstream trend. However, in actual application, due to factors such as long project development cycle, frequent changes in team members, and limitations in development architecture design, existing front-end component resources have not been fully utilized. Moreover, when developing new functions, developers may repeat development work because they do not understand existing reusable components, which not only increases code redundancy, but also prolongs development progress, affecting the final quality and development efficiency of the product, and increasing the overall development cost.

[0003] In summary, how to solve the problem of low development efficiency caused by insufficient utilization of development component resources and limitations of development architecture design while ensuring product quality is an urgent problem that needs to be solved. Summary of the invention

[0004] In view of this, the purpose of the present invention is to provide a development component recommendation method, device, equipment and medium, which can solve the problem of low development efficiency caused by insufficient utilization of development component resources and limitations of development architecture design while ensuring product quality. The specific scheme is as follows:

[0005] In a first aspect, the present application provides a development component recommendation method, comprising:

[0006] Determine the design image data and project text information corresponding to the target development project, and analyze the design image data using a preset image processing model to obtain the project visual features in the design image data;

[0007] Analyzing the project text information using a preset text processing model to obtain a first text feature and a second text feature in the project text information; wherein the first text feature is a feature characterizing project requirement information in the project text information, and the second text feature is a feature characterizing project style information in the project text information;

[0008] Integrate the project visual feature, the first text feature and the second text feature to obtain a corresponding integrated feature, and enhance the integrated feature using a preset feature enhancement algorithm to obtain a first enhanced feature and a second enhanced feature; wherein the distribution of the integrated feature embodied in the first enhanced feature and the second enhanced feature are different from each other;

[0009] Respectively determining a number of component features corresponding to each of the pre-collected development components, and evaluating the development components based on the first enhancement feature, the second enhancement feature, and the number of component features to obtain a number of target recommendation scores corresponding to the development components;

[0010] The development components are sorted according to the target recommendation scores, and development components to be recommended are determined from the development components based on the sorted development component sequence, so as to use the development components to be recommended for development component recommendation for the target development project.

[0011] Optionally, before analyzing the design image data by using a preset image processing model, the method further includes:

[0012] Collecting relevant image data corresponding to historical development projects, and using the relevant image data to construct a model training set;

[0013] The initial model is trained based on the model training set to obtain the preset image processing model; the initial model is a visual transformer model based on the attention mechanism.

[0014] Optionally, the step of integrating the project visual feature, the first text feature, and the second text feature to obtain a corresponding integrated feature, and enhancing the integrated feature using a preset feature enhancement algorithm to obtain a first enhanced feature and a second enhanced feature includes:

[0015] Using a preset splicing method to splice and combine the project visual feature, the first text feature and the second text feature to obtain a corresponding integrated feature;

[0016] Performing a first nonlinear transformation process on the integrated feature to obtain a first enhanced feature, and performing a second nonlinear transformation process on the integrated feature to obtain a second enhanced feature;

[0017] The first nonlinear transformation process and the second nonlinear transformation process are different types of nonlinear transformation processes.

[0018] Optionally, the respectively determining a plurality of component features corresponding to a plurality of development components collected in advance includes:

[0019] Determine respectively the description text information corresponding to each of the pre-collected development components, and input the description text information into the preset text processing model to obtain corresponding component text features;

[0020] Development component vectors of the development components are determined based on the component information of the development components, and the development component vectors are concatenated with the component text features to obtain component features corresponding to the development components.

[0021] Optionally, the evaluating the plurality of development components based on the first enhanced feature, the second enhanced feature, and the plurality of component features to obtain a plurality of target recommendation scores corresponding to the plurality of development components includes:

[0022] respectively calculating dot products between the first enhancement feature and component features corresponding to the plurality of development components to determine a plurality of first recommendation scores corresponding to the plurality of development components;

[0023] respectively calculating dot products between the second enhancement feature and component features corresponding to the plurality of development components to determine a plurality of second recommendation scores corresponding to the plurality of development components;

[0024] A plurality of target recommendation scores corresponding to the plurality of development components are determined based on the first recommendation score and the second recommendation score.

[0025] Optionally, determining a number of target recommendation scores corresponding to the number of development components based on the first recommendation score and the second recommendation score includes:

[0026] Using the sum of the first recommendation score and the second recommendation score as a number of target recommendation scores corresponding to the number of development components;

[0027] Alternatively, a weighted sum is performed on the first recommendation score and the second recommendation score, and the obtained sum is used as the target recommendation scores corresponding to the development components.

[0028] Optionally, the sorting of the plurality of development components according to the plurality of target recommendation scores, and determining the development components to be recommended from the plurality of development components based on the sorted development component sequence, includes:

[0029] sorting the plurality of development components based on the plurality of target recommendation scores to obtain sorted development components;

[0030] A preset number of development components are selected from the sorted development components in descending order of the target recommendation scores, and the preset number of development components are used as development components to be recommended; or, a target development component whose target recommendation score is higher than a preset score threshold among the sorted development components is determined, and the target development component is used as the development component to be recommended.

[0031] In a second aspect, the present application provides a development component recommendation device, comprising:

[0032] A data analysis module, used to determine the design image data and project text information corresponding to the target development project, and analyze the design image data using a preset image processing model to obtain the project visual features in the design image data;

[0033] An information analysis module, used to analyze the project text information using a preset text processing model to obtain a first text feature and a second text feature in the project text information; wherein the first text feature is a feature characterizing project demand information in the project text information, and the second text feature is a feature characterizing project style information in the project text information;

[0034] A feature enhancement module, used for integrating the project visual feature, the first text feature and the second text feature to obtain a corresponding integrated feature, and enhancing the integrated feature using a preset feature enhancement algorithm to obtain a first enhanced feature and a second enhanced feature; wherein the distribution of the integrated feature embodied in the first enhanced feature and the second enhanced feature are different from each other;

[0035] a component evaluation module, configured to respectively determine a plurality of component features corresponding to each of the pre-collected plurality of development components, and evaluate the plurality of development components based on the first enhanced feature, the second enhanced feature, and the plurality of component features to obtain a plurality of target recommendation scores corresponding to the plurality of development components;

[0036] A component determination module is used to sort the several development components according to the several target recommendation scores, and determine the development components to be recommended from the several development components based on the development component order obtained after sorting, so as to use the development components to be recommended to recommend development components for the target development project.

[0037] In a third aspect, the present application provides an electronic device, including:

[0038] Memory, used to store computer programs;

[0039] The processor is used to execute the computer program to implement the aforementioned development component recommendation method.

[0040] In a fourth aspect, the present application provides a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, the aforementioned development component recommendation method is implemented.

[0041] In this embodiment, the design image data and project text information corresponding to the target development project are determined, and the design image data are analyzed using a preset image processing model to obtain the project visual features in the design image data; the project text information is analyzed using a preset text processing model to obtain the first text feature and the second text feature in the project text information; wherein the first text feature is a feature that characterizes the project demand information in the project text information, and the second text feature is a feature that characterizes the project style information in the project text information; the project visual features, the first text features, and the second text features are integrated to obtain corresponding integrated features, and the integrated features are enhanced using a preset feature enhancement algorithm. Performing enhancement processing to obtain a first enhancement feature and a second enhancement feature; wherein the distribution of the integrated features embodied in the first enhancement feature and the second enhancement feature are different from each other; respectively determining a number of component features corresponding to a number of development components collected in advance, and evaluating the number of development components based on the first enhancement feature, the second enhancement feature and the number of component features to obtain a number of target recommendation scores corresponding to the number of development components; sorting the number of development components according to the number of target recommendation scores, and determining a development component to be recommended from the number of development components based on the development component sequence obtained after sorting, so as to use the development component to be recommended to recommend development components for the target development project. As can be seen from the above, the present application first determines the design image data and project text information corresponding to the target development project, and analyzes the design image data and the project text information based on the preset image processing model and the preset text processing model to obtain the corresponding project visual features, the first text features and the second text features, and then integrates the project visual features, the first text features and the second text features, enhances the integrated features obtained after integration, to obtain the first enhanced features and the second enhanced features, and determines the several component characteristics corresponding to the several development components, evaluates the several development components based on the first enhanced features, the second enhanced features and the several component features, to obtain the several target recommendation scores corresponding to them, and determines the development components to be recommended from the several development components based on the several target recommendation scores, so as to recommend the development components for the target development project. In this way, through the above process of the present application, the development components are recommended for the target development project during the development process, the process of selecting and using the development components is optimized, and the intelligent recommendation mechanism is used to help developers quickly find the development components suitable for the current project, reduce unnecessary duplication of work, and thereby solve the problem of low development efficiency caused by insufficient utilization of development component resources and limitations of development architecture design while ensuring product quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0043] Figure 1 A flowchart of a development component recommendation method disclosed in this application;

[0044] Figure 2 A schematic diagram of a system module of a development component recommendation method disclosed in this application;

[0045] Figure 3 A timing diagram of a component feature determination process of a development component disclosed in the present application;

[0046] Figure 4 A flowchart of a specific development component recommendation method disclosed in this application;

[0047] Figure 5 A schematic diagram of a development component recommendation device structure disclosed in this application;

[0048] Figure 6 This is a structural diagram of an electronic device disclosed in this application. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] In the actual application process, due to factors such as long project development cycle, frequent changes in team members and limitations in development architecture design, the existing front-end component resources are not fully utilized. And when developing new functions, developers may repeat the development work because they do not understand the existing reusable components, which not only increases the redundancy of the code, but also prolongs the development progress, affecting the final quality and development efficiency of the product, and increasing the overall development cost.

[0051] In order to overcome the above technical problems, the present application provides a development component recommendation method to solve the problem of low development efficiency caused by insufficient utilization of development component resources and limitations of development architecture design while ensuring product quality.

[0052] See also Figure 1 As shown, an embodiment of the present invention discloses a development component recommendation method, including:

[0053] Step S11, determining the design image data and project text information corresponding to the target development project, and analyzing the design image data using a preset image processing model to obtain project visual features in the design image data.

[0054] In this embodiment, the design image data and project text information corresponding to the target development project are first determined, and the design image data are analyzed using a preset image processing model to obtain the project visual features in the design image data. The target development project is a development project that requires development component recommendations based on the needs of developers; the design image data can be a visual design page image of the target development project, that is, a prototype design drawing of the target development project; the project visual features are feature values ​​that can reflect the layout characteristics of the target development project; the preset image processing model can capture and express the key features of the development layout of the target development project, and can be a model obtained by training the initial model based on a preset algorithm. The preset image processing model can also be a pre-trained model trained on a large-scale image data set, and the pre-trained model includes but is not limited to VGG (Visual Geometry Group), ResNet (Residual Neural Network), etc.

[0055] It can be understood that before using the preset image processing model to analyze the design image data, it is necessary to first train the preset image processing model, and the processing flow is as follows: collect the corresponding relevant image data from the historical development project, and use the relevant image data to build a model training set; train the initial model based on the model training set to obtain the preset image processing model; the initial model is a visual transformer model based on the attention mechanism. Among them, the historical development project can be a relevant development project collected from the development community, or a local relevant development project provided by the developer. That is, the model is trained using the relevant image data in the historical development project, the corresponding relevant image data is collected from the historical development project, and the relevant image data is subjected to data enhancement processing such as random cropping, rotation, flipping, etc. to increase the diversity of the training data, and a model training set is built based on the relevant image data, and the constructed visual transformer (i.e., VisionTransformer, ViT) model based on the attention mechanism is trained using the model training set to obtain the preset image processing model output after training. Regarding the construction of the visual transformer model, specifically, this embodiment can first determine the architecture of the visual transformer, embed the image, construct a transformer structure (a deep learning model architecture) containing multiple encoder layers, and add a multi-layer perceptron as a classification head on its output to realize the construction of the visual transformer model. It can be understood that in addition to obtaining the project visual features in the design image data by using the preset image processing model, this embodiment can also obtain the project visual features in the design image data by using an algorithm for extracting image features, wherein the algorithm includes but is not limited to SIFT (i.e., Scale-Invariant Feature Transform), SURF (i.e., Speeded-Up Robust Features), corner detection algorithm, etc. After obtaining the preset image processing model, this embodiment can input the design image data of the target development project into the preset image processing model so that the preset image processing model extracts the features of the design image data and obtains the project visual features output by the preset image processing model. For example, assuming that the target development project is the development of web page A, and the original image data is a web page design drawing P. First, the web design image P needs to be preprocessed, such as converted into a digital form, to ensure that its format meets the requirements of the ViT model. The processed image data is then input into the trained ViT model, which analyzes the image data and extracts features that reflect the user interface layout and style of the web design image P, and regards these features as project visual features.

[0056] It should be pointed out that if Figure 2 The figure shows a system module diagram of a development component recommendation method provided by the present application. The development component recommendation method of the present application is applicable to various electronic devices, such as smart phones, personal computers, tablet computers, etc., and is applicable to project development scenarios that require rapid iteration and high-quality completion, and is particularly applicable to complex software development scenarios, especially those that require rapid response to market changes and improved user experience. The process of the method includes six modules: data acquisition module, visual feature extraction module, text feature extraction module, feature fusion module, component feature determination module, and recommendation result determination module. Among them, the data acquisition module is responsible for collecting various data required for the development of the target project, such as design image data, project text information and available development component sets; the visual feature extraction module is used to extract project visual features from the design image data, which can help identify the design style and layout requirements of the project; the text feature extraction module can extract two different text features from the project text information respectively, so as to understand the project requirements more comprehensively; the feature fusion module can integrate the project visual features with the two text features, and generate two different enhanced features through a specific algorithm; the component feature determination module is used to analyze each component in the development component set and determine its respective component features; the recommendation result determination module is responsible for selecting the most suitable component for the target project from the development component set based on the enhanced features and component features generated by the above modules as the recommendation result. In this way, the present embodiment first obtains the development data of the target development project, and extracts the relevant project visual features from the development data, which can help identify the design style and layout requirements of the project, understand the visual elements such as the layout and color matching of the project, effectively extract the project visual features that are valuable to the project development from the design image data, and provide an important reference basis for the subsequent development component recommendation, thereby improving the selection efficiency of the development component, ensuring that the selected component is highly matched with the project layout requirements, and speeding up the development speed.

[0057] Step S12: Analyze the project text information using a preset text processing model to obtain a first text feature and a second text feature in the project text information; wherein the first text feature is a feature that characterizes the project requirement information in the project text information, and the second text feature is a feature that characterizes the project style information in the project text information.

[0058] In this embodiment, the project text information is analyzed using a preset text processing model to obtain a first text feature and a second text feature in the project text information. The first text feature is a feature that describes the project requirement information in the project text information; the second text feature is a feature that characterizes the project style information in the project text information; the project text information contains requirement description information and style description information about the target development project, that is, the project requirement information and the project style information; the project requirement information includes but is not limited to requirement description text information, functional requirement text information, user feedback text information, etc.; the project style information includes but is not limited to architecture design style information, color design style information, interaction design style information, etc.

[0059] Specifically, this embodiment can first determine the project requirement information and the project style information in the project text information, and then perform preprocessing such as word segmentation on the project requirement information and the project style information to convert the project requirement information and the project style information into a computer-processable form, and input the preprocessed project requirement information into the preset text processing model so that the preset text processing model can understand and extract the semantic features in the project requirement information and the relationship features between the text segmentation to obtain the first text feature representing the project requirement information. Similarly, the preprocessed project style information is input into the preset text processing model so that the preset text processing model can extract the semantic features of the segmentation of the project style information and the relationship features therebetween to obtain the second text feature representing the project style information. It should be noted that the preset text processing model can be a model such as SBERT (i.e., Sentence-BERT, a semantic similarity model), RNN (i.e., Recurrent Neural Network), LSTM (i.e., Long Short-term Memory Networks), etc. In this way, this embodiment uses a preset text processing model to analyze the project text information to obtain a first text feature about the project requirement information and a second text feature about the project style information in the project text information, so as to more comprehensively understand the project requirements and improve the efficiency of selecting development components. By subsequently evaluating the components using the first text feature and the second text feature, it is ensured that the selected components are highly matched with the overall project requirements, thereby accelerating the development speed.

[0060] Step S13: Integrate the project visual features, the first text features and the second text features to obtain corresponding integrated features, and enhance the integrated features using a preset feature enhancement algorithm to obtain first enhanced features and second enhanced features; wherein the distribution of the integrated features embodied in the first enhanced feature and the second enhanced feature are different from each other.

[0061] In this embodiment, the project visual feature, the first text feature and the second text feature are combined to generate a corresponding comprehensive integrated feature, and the integrated feature is enhanced using a preset feature enhancement algorithm to obtain a first enhanced feature and a second enhanced feature. The preset feature enhancement algorithm may be a preset nonlinear transformation method; the distribution of the integrated feature embodied by the first enhanced feature is different from the distribution of the integrated feature embodied by the second enhanced feature.

[0062] Specifically, this embodiment can use a preset splicing method to splice and combine the project visual features, the first text features, and the second text features to obtain a corresponding integrated feature; perform a first nonlinear transformation process on the integrated feature to obtain a first enhanced feature, and perform a second nonlinear transformation process on the integrated feature to obtain a second enhanced feature; wherein the first nonlinear transformation process and the second nonlinear transformation process are different types of nonlinear transformation processes. That is, this embodiment can splice and combine the project visual features, the first text features, and the second text features based on a preset splicing method to form a comprehensive integrated feature, and perform two nonlinear transformation processes on the integrated feature, namely the first nonlinear transformation process and the second nonlinear transformation process, to generate two different enhanced features, namely the first enhanced feature and the second enhanced feature. Among them, the preset splicing method can be horizontal splicing or vertical splicing; the enhanced feature is a higher-dimensional feature that reflects the interaction between the three features of the project visual feature, the first text feature and the second text feature, and is used to enhance the interaction between features to obtain richer feature representation; the first enhanced feature is a feature obtained based on the first nonlinear transformation processing, which is used to reflect the feature interaction from one perspective; the second enhanced feature is a feature obtained based on the second nonlinear transformation processing, which is used to reflect the feature interaction from another perspective; the nonlinear transformation processing method includes but is not limited to Log (logarithmic) transformation processing, Box-Cox transformation processing (a generalized power transformation method), hyperbolic tangent transformation processing, etc., which can improve the distribution of the enhanced features and ensure the effectiveness and accuracy of the features. It should be pointed out that in the process of splicing and combining the project visual features, the first text features and the second text features, the present embodiment can also give different weights according to the importance of each feature to the target development project, that is, based on the importance of the target development project, the project visual features, the first text features and the second text features are quantified respectively, and then the project visual features, the first text features and the second text features are weighted based on their respective quantification results. This method can improve the accuracy and generalization ability of component evaluation and achieve more personalized component recommendation services. In this way, the present embodiment integrates the project visual features, the first text features and the second text features, and processes the integrated features through a specific algorithm to generate two different enhanced features, comprehensively consider multiple factors, and perform high-level fusion processing to serve the subsequent target project analysis or decision support, and improve the accuracy of component recommendation.

[0063] Step S14, respectively determine a number of component features corresponding to each of the pre-collected development components, and evaluate the development components based on the first enhancement feature, the second enhancement feature and the number of component features to obtain a number of target recommendation scores corresponding to the development components.

[0064] In this embodiment, the component features corresponding to each development component in the pre-collected development components are determined, and the development components are evaluated based on the first enhanced features, the second enhanced features and the component features, and the target recommendation scores corresponding to each of the development components are obtained by calculation. Among them, the component features can be the technical characteristics, usage scenarios, etc. of the development components; the development components can be components obtained from the open source community or customized according to the specific needs of the target development project, and a development component collection is constructed based on the development components and their corresponding descriptive text information for centralized storage; the development component collection contains at least one development component used to develop the target development project and its detailed descriptive text information; the target recommendation score is a reference indicator used to characterize whether the corresponding development component is worthy of being recommended for use in the current target development project. The higher the target recommendation score, the more worthy of recommendation the corresponding development component is.

[0065] It should be noted that in order to more comprehensively obtain the component features of the several development components, the present embodiment can determine the component features of the several development components based on multiple aspects, and the processing flow is as follows: respectively determine the descriptive text information corresponding to the several development components collected in advance, and input the descriptive text information into the preset text processing model to obtain the corresponding component text features; determine the development component vectors of the several development components based on the component information of the several development components, and splice the development component vectors with the component text features to obtain several component features corresponding to the several development components. Among them, the component information is used to characterize the properties, structure, content and behavior information of the development component, including but not limited to the component's property information such as color design, position information, structural information such as position in the page structure, relationship with other development components, content information such as text content, image content, and behavior information such as interactive behavior; the descriptive text information is used to record the name, function, appearance design, interactive behavior and other contents of the development component. For example, for a web page navigation bar component, its descriptive text information may include its name such as the top navigation bar, functions such as providing internal link navigation of the website, styles such as background color and font size, and interactive effects such as dynamic changes when the mouse hovers over the component. That is, this embodiment first determines the descriptive text information corresponding to each of the several development components collected in advance, and inputs the descriptive text information into the preset text processing model to obtain the component text features output by the preset text processing model, and at the same time determines the development component vectors of the several development components based on the component information of the several development components, and splices the development component vectors with the component text features to obtain several component features corresponding to the several development components. Figure 3 Shown is a timing diagram of a component feature determination process of a development component provided by the present application. It should be noted that if the target development project is a front-end development project, the several development components may be front-end components, including but not limited to the navigation bar of a web page, a user homepage, etc. It should be further noted that the present embodiment can evaluate the several development components by calculating the similarity or matching degree between the first enhanced feature and the component feature, and between the second enhanced feature and the component feature, so as to ensure the relevance and practicality of the recommendation results. In this way, the present embodiment determines the component features of the development component based on the descriptive text information and component information of the development component, and evaluates several development components based on the unique component features, the first enhanced features, and the second enhanced features determined comprehensively to determine the corresponding several target recommendation scores, which not only considers the visual and textual features of the project, but also fully considers the characteristics of the component itself, thereby ensuring the relevance and practicality of the recommendation results.

[0066] Step S15, sorting the several development components according to the several target recommendation scores, and determining the development components to be recommended from the several development components based on the sorted development component order, so as to use the development components to be recommended for development component recommendation for the target development project.

[0067] In this embodiment, the development components are sorted according to the target recommendation scores, so as to determine the development components to be recommended from the development components based on the development component order obtained after sorting, and the development components to be recommended are recommended for the target development project using the determined development components to be recommended. The development components to be recommended are development components that are determined to be worthy of recommendation after comparison of the target recommendation scores.

[0068] Specifically, this embodiment can sort the several development components based on the several target recommendation scores to obtain sorted development components; select a preset number of development components from the sorted development components in the order of the target recommendation scores from high to low, so as to use the preset number of development components as the development components to be recommended, or determine the target development components whose target recommendation scores are higher than the preset score threshold in the sorted development components, and use the target development components as the development components to be recommended. That is, after sorting the several development components based on the several target recommendation scores, this embodiment can select the development components to be recommended according to the sorted development components. There are two methods for selecting the development components to be recommended. In a specific implementation, this embodiment can select the development components with higher target recommendation scores than other development components from the arranged development components based on the determined preset number as the development components to be recommended. That is, select a preset number of development components from the sorted development components in the order of the target recommendation scores from high to low. The preset number can be a default preset number, or a number set by the developer through the quantity setting interface based on his own development needs. In another specific implementation, this embodiment can set a preset score threshold, compare the target recommendation scores of the development components in the sorted development components in order of scores from high to low based on the preset score threshold, determine the position of the first development component below the preset score threshold, and use the development component compared before the current position as the development component to be recommended. In this way, this embodiment recommends development components based on the target recommendation score determined by comprehensively considering the characteristics of the project and the characteristics of the development component itself, which can significantly improve the efficiency and quality of the development project, reduce development costs, and also provide developers with a scientific and reasonable basis for component selection.

[0069] As can be seen from the above, the embodiment of the present application first determines the design image data and project text information corresponding to the target development project, and analyzes the design image data and the project text information based on a preset image processing model and a preset text processing model, respectively, to obtain corresponding project visual features, first text features and second text features, and then integrates the project visual features, the first text features and the second text features, enhances the integrated features obtained after the integration to obtain first enhanced features and second enhanced features, and determines a number of component characteristics corresponding to a number of development components, evaluates the several development components based on the first enhanced features, the second enhanced features and the several component characteristics to obtain their corresponding several target recommendation scores, and determines the development components to be recommended from the several development components based on the several target recommendation scores to recommend development components for the target development project.In this way, through the above process of the embodiment of the present application, on the one hand, the development data of the target development project is obtained, and the relevant project visual features are extracted from the development data, which can help identify the design style and layout requirements of the project, understand the visual elements such as the layout and color matching of the project, and effectively extract the project visual features that are valuable to the project development from the design image data, and provide an important reference basis for the subsequent recommendation of development components; on the one hand, the project text information is analyzed using a preset text processing model to obtain the first text feature about the project requirement information and the second text feature about the project style information in the project text information, so as to more comprehensively understand the project requirements and improve the selection efficiency of the development components. By subsequently using the first text feature and the second text feature to evaluate the components, it is ensured that the selected components are highly matched with the overall project requirements, thereby speeding up the development; on the one hand, the project visual features, the first text features and the second text features are integrated, and the integrated features are processed by a specific algorithm. The first enhancement feature is used to process the features and generate two different enhanced features. A variety of factors are comprehensively considered to perform high-level fusion processing to serve the subsequent target project analysis or decision support and improve the accuracy of component recommendation. On the one hand, several development components are evaluated based on the comprehensively determined unique component features, the first enhancement feature and the second enhancement feature to determine their corresponding target recommendation scores. Not only the visual and text features of the project are considered, but also the characteristics of the component itself are fully considered, thereby ensuring the relevance and practicality of the recommendation results. On the other hand, by extracting multi-dimensional features of the original image data and original text information of the target project and combining the component features of the development components, accurate recommendation of development components is achieved, the process of selecting and using development components is optimized, and an intelligent recommendation mechanism is used to help developers quickly find development components suitable for the current project, reduce unnecessary repetitive work, and thereby solve the problem of low development efficiency caused by insufficient utilization of development component resources and limitations of development architecture design while ensuring product quality.

[0070] Based on the above embodiments, it can be seen that through the method of the present application, when evaluating several development components, the evaluation can be performed based on the first enhanced feature and the second enhanced feature respectively, so as to perform a multi-dimensional development component evaluation, thereby effectively improving the development efficiency of the project and the quality of the product. For this reason, this embodiment describes in detail how to evaluate several development components. Figure 4 As shown, an embodiment of the present invention discloses a development component recommendation method, including:

[0071] Step S21: determine the design image data and project text information corresponding to the target development project, and analyze the design image data using a preset image processing model to obtain project visual features in the design image data.

[0072] Step S22: Analyze the project text information using a preset text processing model to obtain a first text feature and a second text feature in the project text information; wherein the first text feature is a feature that characterizes the project requirement information in the project text information, and the second text feature is a feature that characterizes the project style information in the project text information.

[0073] Step S23: Integrate the project visual features, the first text features and the second text features to obtain corresponding integrated features, and enhance the integrated features using a preset feature enhancement algorithm to obtain first enhanced features and second enhanced features; wherein the distribution of the integrated features embodied in the first enhanced feature and the second enhanced feature are different from each other.

[0074] Step S24: respectively determine a plurality of component features corresponding to the pre-collected plurality of development components, and respectively calculate dot products between the first enhancement feature and the component features corresponding to the plurality of development components to determine a plurality of first recommendation scores corresponding to the plurality of development components.

[0075] In this embodiment, several component features corresponding to several development components collected in advance are first determined respectively, and the dot product between the first enhanced feature and the component features corresponding to the several development components is calculated respectively, and the dot product calculated based on the first enhanced feature is used as the several first recommendation scores corresponding to the several development components. Specifically, this embodiment can process each of the several development components, multiply the first enhanced feature with the elements at the corresponding positions of the component features of each development component one by one to form a first product vector, and then add all the element values ​​in the first product vector, and use the obtained sum as the first recommendation score of the development component. In this way, this embodiment evaluates several development components from the dimension of the first enhanced feature by calculating the dot product of the first enhanced feature and the component feature to determine the first recommendation score under the current dimension, which is conducive to combining the characteristics of the development components to achieve efficient and accurate recommendation of the development components, thereby significantly improving the development efficiency and product quality of Internet projects.

[0076] Step S25: Calculate dot products between the second enhancement feature and component features corresponding to the plurality of development components respectively to determine a plurality of second recommendation scores corresponding to the plurality of development components.

[0077] In this embodiment, the dot products between the second enhanced feature and the component features corresponding to the several development components are calculated respectively, and the dot products calculated based on the second enhanced feature are used as the several second recommendation scores corresponding to the several development components. Specifically, this embodiment can process each of the several development components, multiply the second enhanced feature with the elements at the corresponding positions of the component features of each development component one by one to form a second product vector, then add all the element values ​​in the second product vector, and use the obtained sum as the second recommendation score of the development component. In this way, this embodiment evaluates several development components from the dimension of the second enhanced feature by calculating the dot product of the second enhanced feature and the component feature to determine the second recommendation score under the current dimension, which is conducive to combining the characteristics of the development components to achieve efficient and accurate recommendation of the development components, thereby significantly improving the development efficiency and product quality of Internet projects.

[0078] Step S26: Determine a plurality of target recommendation scores corresponding to the plurality of development components based on the first recommendation score and the second recommendation score.

[0079] In this embodiment, the first recommendation score and the second recommendation score are used to determine the target recommendation scores corresponding to the development components. That is, the target recommendation scores of the development components are determined based on the first recommendation score and the second recommendation score.

[0080] It is understandable that in order to determine the target recommendation scores of the several development components, the present embodiment may adopt different determination methods. In a specific implementation, the present embodiment may use the sum of the first recommendation score and the second recommendation score as the target recommendation scores corresponding to the several development components. That is, for each development component, the present embodiment may simply add the first recommendation score and the second recommendation score, and use the total sum obtained as the target recommendation score of the several development components. In another specific implementation, the present embodiment may perform a weighted summation of the first recommendation score and the second recommendation score, and use the obtained summation as the target recommendation scores corresponding to the several development components. That is, for each development component, the present embodiment may perform a weighted summation of the first recommendation score and the second recommendation score according to the actual target development project, and use the obtained summation as the target recommendation scores corresponding to the several development components. It should be noted that the two above-mentioned determination methods can be selected and determined from the preset determination interface based on the recommendation requirements of the developer. In this way, this embodiment determines several target recommendation scores corresponding to several development components based on the first recommendation score and the second recommendation score through two determination methods, thereby enabling developers to choose a determination method based on their own needs. It can effectively solve the problems encountered by developers when selecting reusable components, and ensure that the recommended development components are highly matched with the needs of the target project, thereby improving development efficiency and reducing costs.

[0081] Step S27, sorting the several development components according to the several target recommendation scores, and determining the development components to be recommended from the several development components based on the development component order obtained after sorting, so as to use the development components to be recommended to recommend development components for the target development project.

[0082] Among them, the specific implementation process of steps S21, S22, S23 and S27 can refer to the contents of the aforementioned embodiments, and will not be repeated here.

[0083] As can be seen from the above, the embodiment of the present application first determines the design image data and project text information corresponding to the target development project, and analyzes the design image data and the project text information based on a preset image processing model and a preset text processing model, respectively, to obtain the corresponding project visual features, first text features and second text features, and then integrates the project visual features, the first text features and the second text features, enhances the integrated features obtained after the integration, to obtain the first enhanced features and the second enhanced features, and determines a number of component characteristics corresponding to a number of development components, and calculates the dot product between the first enhanced feature and the component characteristics corresponding to the several development components, and the dot product between the second enhanced feature and the component characteristics, respectively, to obtain the corresponding several first recommendation scores and several second recommendation scores, and determines several target recommendation scores based on the first recommendation scores and the second recommendation scores, and determines the development components to be recommended from the several development components based on the several target recommendation scores, so as to recommend development components for the target development project. In this way, through the above process of the embodiment of the present application, on the one hand, when evaluating several development components, the evaluation is performed based on the first enhanced feature and the second enhanced feature respectively to determine the corresponding first recommendation score and the second recommendation score, which is conducive to combining the characteristics of the development components to achieve efficient and accurate recommendation of the development components, thereby significantly improving the development efficiency and product quality of Internet projects; on the other hand, the target recommendation score can be obtained by weighted summation, which can better reflect the degree of matching between the development component and the target project, so as to perform multi-dimensional development component evaluation, thereby effectively improving the project development efficiency and product quality.

[0084] Accordingly, see Figure 5 As shown, the embodiment of the present application also provides a development component recommendation device, including:

[0085] The data analysis module 11 is used to determine the design image data and project text information corresponding to the target development project, and analyze the design image data using a preset image processing model to obtain the project visual features in the design image data;

[0086] The information analysis module 12 is used to analyze the project text information using a preset text processing model to obtain a first text feature and a second text feature in the project text information; wherein the first text feature is a feature characterizing the project requirement information in the project text information, and the second text feature is a feature characterizing the project style information in the project text information;

[0087] The feature enhancement module 13 is used to integrate the project visual feature, the first text feature and the second text feature to obtain a corresponding integrated feature, and enhance the integrated feature using a preset feature enhancement algorithm to obtain a first enhanced feature and a second enhanced feature; wherein the distribution of the integrated feature embodied in the first enhanced feature and the second enhanced feature are different from each other;

[0088] A component evaluation module 14 is used to respectively determine a number of component features corresponding to each of the pre-collected development components, and evaluate the development components based on the first enhanced feature, the second enhanced feature and the number of component features to obtain a number of target recommendation scores corresponding to the development components;

[0089] The component determination module 15 is used to sort the several development components according to the several target recommendation scores, and determine the development components to be recommended from the several development components based on the development component order obtained after sorting, so as to use the development components to be recommended to recommend development components for the target development project.

[0090] As can be seen from the above, the embodiment of the present application first determines the design image data and project text information corresponding to the target development project, and analyzes the design image data and the project text information based on a preset image processing model and a preset text processing model, respectively, to obtain corresponding project visual features, first text features and second text features, and then integrates the project visual features, the first text features and the second text features, enhances the integrated features obtained after the integration to obtain first enhanced features and second enhanced features, and determines a number of component characteristics corresponding to a number of development components, evaluates the several development components based on the first enhanced features, the second enhanced features and the several component characteristics to obtain their corresponding several target recommendation scores, and determines the development components to be recommended from the several development components based on the several target recommendation scores to recommend development components for the target development project. In this way, through the above-mentioned process of the embodiment of the present application, development components are recommended for the target development project during the development process, the process of selecting and using development components is optimized, and an intelligent recommendation mechanism is used to help developers quickly find development components suitable for the current project, thereby reducing unnecessary repetitive work, thereby ensuring product quality while solving the problem of low development efficiency caused by insufficient utilization of development component resources and limitations of development architecture design.

[0091] In some specific implementations, the development component recommendation device may further include:

[0092] A training set construction unit, used to collect relevant image data corresponding to historical development projects and construct a model training set using the relevant image data;

[0093] A model training unit is used to train an initial model based on the model training set to obtain the preset image processing model; the initial model is a visual transformer model based on an attention mechanism.

[0094] In some specific implementations, the feature enhancement module 13 may specifically include:

[0095] A feature combination unit, used for combining the project visual feature, the first text feature and the second text feature by using a preset combination method to obtain a corresponding integrated feature;

[0096] a feature processing unit, configured to perform a first nonlinear transformation process on the integrated feature to obtain a first enhanced feature, and perform a second nonlinear transformation process on the integrated feature to obtain a second enhanced feature;

[0097] The first nonlinear transformation process and the second nonlinear transformation process are different types of nonlinear transformation processes.

[0098] In some specific implementations, the component evaluation module 14 may specifically include:

[0099] A model input unit, used to respectively determine the description text information corresponding to each of the pre-collected development components, and input the description text information into the preset text processing model to obtain corresponding component text features;

[0100] A vector splicing unit is used to determine development component vectors of the several development components based on the component information of the several development components, and splice the development component vectors with the component text features to obtain several component features corresponding to the several development components.

[0101] In some specific implementations, the component evaluation module 14 may specifically include:

[0102] A first dot product calculation submodule, used to calculate dot products between the first enhancement feature and component features corresponding to the plurality of development components, respectively, to determine a plurality of first recommendation scores corresponding to the plurality of development components;

[0103] A second dot product calculation submodule, used to calculate the dot products between the second enhanced feature and the component features corresponding to the plurality of development components, respectively, to determine a plurality of second recommendation scores corresponding to the plurality of development components;

[0104] The score determination submodule is used to determine a plurality of target recommendation scores corresponding to the plurality of development components based on the first recommendation score and the second recommendation score.

[0105] In some specific implementations, the score determination submodule may specifically include:

[0106] A first score determination unit, configured to use a sum of the first recommendation score and the second recommendation score as a plurality of target recommendation scores corresponding to the plurality of development components;

[0107] Or, a second score determination unit is used to perform a weighted sum of the first recommendation score and the second recommendation score, and use the obtained sum as the multiple target recommendation scores corresponding to the multiple development components.

[0108] In some specific implementations, the component determination module 15 may specifically include:

[0109] A component sorting unit, configured to sort the plurality of development components based on the plurality of target recommendation scores to obtain sorted development components;

[0110] A component determination unit is used to select a preset number of development components from the sorted development components in descending order of the target recommendation scores, so as to use the preset number of development components as development components to be recommended, or to determine a target development component whose target recommendation score is higher than a preset score threshold among the sorted development components, and use the target development component as the development component to be recommended.

[0111] Furthermore, the present application also discloses an electronic device. Figure 6 It is a structural diagram of an electronic device 20 shown according to an exemplary embodiment, and the content in the figure cannot be regarded as any limitation on the scope of use of this application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input and output interface 25 and a communication bus 26. Among them, the memory 22 is used to store a computer program, and the computer program is loaded and executed by the processor 21 to implement the relevant steps in the development component recommendation method disclosed in any of the aforementioned embodiments. In addition, the electronic device 20 in this embodiment can specifically be an electronic computer.

[0112] In this embodiment, the power supply 23 is used to provide working voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and the external device, and the communication protocol it follows is any communication protocol that can be applied to the technical solution of the present application, and is not specifically limited here; the input and output interface 25 is used to obtain external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0113] In addition, the memory 22 as a carrier for resource storage may be a read-only memory, a random access memory, a disk or an optical disk, etc. The resources stored thereon may include an operating system 221, a computer program 222, etc., and the storage method may be temporary storage or permanent storage.

[0114] The operating system 221 is used to manage and control the hardware devices on the electronic device 20 and the computer program 222, which can be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program that can be used to complete the development component recommendation method performed by the electronic device 20 disclosed in any of the aforementioned embodiments, the computer program 222 can further include a computer program that can be used to complete other specific tasks.

[0115] Furthermore, the present application also discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, the aforementioned disclosed development component recommendation method is implemented. For the specific steps of the method, reference may be made to the corresponding contents disclosed in the aforementioned embodiments, and no further description will be given here.

[0116] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0117] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0118] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0119] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0120] The technical solution provided by the present application is introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for general technicians in this field, according to the idea of ​​the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A development component recommendation method, characterized in that: include: Determine the design image data and project text information corresponding to the target development project, and analyze the design image data using a preset image processing model to obtain the project visual features in the design image data; Analyzing the project text information using a preset text processing model to obtain a first text feature and a second text feature in the project text information; wherein the first text feature is a feature characterizing project requirement information in the project text information, and the second text feature is a feature characterizing project style information in the project text information; Integrate the project visual feature, the first text feature and the second text feature to obtain a corresponding integrated feature, and enhance the integrated feature using a preset feature enhancement algorithm to obtain a first enhanced feature and a second enhanced feature; wherein the distribution of the integrated feature embodied in the first enhanced feature and the second enhanced feature are different from each other; Respectively determining a number of component features corresponding to each of the pre-collected development components, and evaluating the development components based on the first enhancement feature, the second enhancement feature, and the number of component features to obtain a number of target recommendation scores corresponding to the development components; The development components are sorted according to the target recommendation scores, and development components to be recommended are determined from the development components based on the sorted development component sequence, so as to use the development components to be recommended for development component recommendation for the target development project.

2. The development component recommendation method according to claim 1, characterized in that: Before analyzing the design image data by using a preset image processing model, the method further includes: Collecting relevant image data corresponding to historical development projects, and using the relevant image data to construct a model training set; The initial model is trained based on the model training set to obtain the preset image processing model; the initial model is a visual transformer model based on the attention mechanism.

3. The development component recommendation method according to claim 1, characterized in that: The step of integrating the project visual feature, the first text feature, and the second text feature to obtain a corresponding integrated feature, and enhancing the integrated feature using a preset feature enhancement algorithm to obtain a first enhanced feature and a second enhanced feature includes: Using a preset splicing method to splice and combine the project visual feature, the first text feature and the second text feature to obtain a corresponding integrated feature; Performing a first nonlinear transformation process on the integrated feature to obtain a first enhanced feature, and performing a second nonlinear transformation process on the integrated feature to obtain a second enhanced feature; The first nonlinear transformation process and the second nonlinear transformation process are different types of nonlinear transformation processes.

4. The development component recommendation method according to claim 1, characterized in that: The step of respectively determining the component features corresponding to the pre-collected development components includes: Determine respectively the description text information corresponding to each of the pre-collected development components, and input the description text information into the preset text processing model to obtain corresponding component text features; Development component vectors of the development components are determined based on the component information of the development components, and the development component vectors are concatenated with the component text features to obtain component features corresponding to the development components.

5. The development component recommendation method according to claim 1, characterized in that: The evaluating the plurality of development components based on the first enhanced feature, the second enhanced feature, and the plurality of component features to obtain a plurality of target recommendation scores corresponding to the plurality of development components includes: respectively calculating dot products between the first enhancement feature and component features corresponding to the plurality of development components to determine a plurality of first recommendation scores corresponding to the plurality of development components; respectively calculating dot products between the second enhancement feature and component features corresponding to the plurality of development components to determine a plurality of second recommendation scores corresponding to the plurality of development components; A plurality of target recommendation scores corresponding to the plurality of development components are determined based on the first recommendation score and the second recommendation score.

6. The development component recommendation method according to claim 5, characterized in that: The determining a plurality of target recommendation scores corresponding to the plurality of development components based on the first recommendation score and the second recommendation score includes: Using the sum of the first recommendation score and the second recommendation score as a number of target recommendation scores corresponding to the number of development components; Alternatively, a weighted sum is performed on the first recommendation score and the second recommendation score, and the obtained sum is used as the target recommendation scores corresponding to the development components.

7. The development component recommendation method according to any one of claims 1 to 6, characterized in that: The step of sorting the plurality of development components according to the plurality of target recommendation scores, and determining the development components to be recommended from the plurality of development components based on the sorted development component sequence, includes: sorting the plurality of development components based on the plurality of target recommendation scores to obtain sorted development components; A preset number of development components are selected from the sorted development components in descending order of the target recommendation scores, and the preset number of development components are used as development components to be recommended; or, a target development component whose target recommendation score is higher than a preset score threshold among the sorted development components is determined, and the target development component is used as the development component to be recommended.

8. A development component recommendation device, characterized in that: include: A data analysis module, used to determine the design image data and project text information corresponding to the target development project, and analyze the design image data using a preset image processing model to obtain the project visual features in the design image data; An information analysis module, used to analyze the project text information using a preset text processing model to obtain a first text feature and a second text feature in the project text information; wherein the first text feature is a feature characterizing project demand information in the project text information, and the second text feature is a feature characterizing project style information in the project text information; A feature enhancement module, used for integrating the project visual feature, the first text feature and the second text feature to obtain a corresponding integrated feature, and enhancing the integrated feature using a preset feature enhancement algorithm to obtain a first enhanced feature and a second enhanced feature; wherein the distribution of the integrated feature embodied in the first enhanced feature and the second enhanced feature are different from each other; a component evaluation module, configured to respectively determine a plurality of component features corresponding to each of the pre-collected plurality of development components, and evaluate the plurality of development components based on the first enhanced feature, the second enhanced feature, and the plurality of component features to obtain a plurality of target recommendation scores corresponding to the plurality of development components; A component determination module is used to sort the several development components according to the several target recommendation scores, and determine the development components to be recommended from the several development components based on the development component order obtained after sorting, so as to use the development components to be recommended to recommend development components for the target development project.

9. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the development component recommendation method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: Used to store a computer program; wherein, when the computer program is executed by a processor, the development component recommendation method as described in any one of claims 1 to 7 is implemented.

Images