Micro-front-end supply chain platform collaboration method and system based on multi-team development

Abstract

The invention relates to the technical field of development collaboration, and relates to a micro-front-end supply chain platform collaboration method and system based on multi-team development, and the method comprises the steps: carrying out the function analysis of a to-be-developed supply chain scheme, obtaining a platform function set, carrying out the function feature matching and development specification of a plurality of teams according to the platform function set, and obtaining a platform function set; standard development information is obtained; performing multi-team module development and standard review on the platform function set according to the standard development information to obtain a standard code set; performing isolation deployment and micro-application integration on the standard code set to obtain a supply chain front end; acquiring a user behavior log by using the supply chain front end, and performing behavior feature extraction and behavior analysis on the user behavior log to obtain an analyzed user behavior; and carrying out dynamic collaborative loading on the front end of the supply chain according to the analyzed user behavior. According to the invention, the efficiency of micro-front-end development collaboration can be improved.

Description

Technical Field

[0001] The present invention relates to the field of development collaboration technology, and in particular to a micro-front-end supply chain platform collaboration method and system based on multi-team development. Background Art

[0002] The supply chain finance platform is a financial service platform that aims to provide financing, settlement, risk management and other services to all participants in the supply chain. Micro-front-end development is a front-end architecture model that aims to decompose a single front-end application into smaller, more independent components or micro-applications to better manage and maintain complex applications. The micro-front-end supply chain platform is a supply chain finance platform created using micro-front-end development technology.

[0003] The existing collaboration methods of micro-frontend supply chain platforms are mostly based on team coordination. By coordinating multiple teams, a micro-frontend supply chain platform is uniformly developed and unified status management is performed. In actual applications, the collaboration methods based on team coordination may have problems such as low collaboration efficiency, low platform scalability, and complex status management, which may lead to low efficiency in micro-frontend development collaboration. Summary of the invention

[0004] The present invention provides a micro-front-end supply chain platform collaboration method and system based on multi-team development, the main purpose of which is to solve the problem of low efficiency in micro-front-end development collaboration.

[0005] To achieve the above objectives, the present invention provides a micro-frontend supply chain platform collaboration method based on multi-team development, including:

[0006] Perform functional analysis on the supply chain solution to be developed to obtain the platform function set, and perform functional feature matching and development specifications on multiple teams based on the platform function set to obtain standardized development information;

[0007] Perform multi-team module development and specification review on the platform function set according to the specification development information to obtain a specification code set;

[0008] The standard code set is deployed in isolation and micro-applications are integrated to obtain a supply chain front end;

[0009] Utilizing the supply chain front end to obtain user behavior logs, extracting behavior features and analyzing behavior on the user behavior logs, and obtaining analyzed user behaviors;

[0010] The supply chain front end is dynamically and collaboratively loaded based on the analyzed user behavior.

[0011] Optionally, the function analysis of the supply chain solution to be developed is performed to obtain a platform function set, including:

[0012] Perform text denoising, word segmentation and word vector encoding on the supply chain solution to be developed to obtain the solution word feature set;

[0013] Performing topic modeling on the scheme word feature set to obtain a topic word feature group;

[0014] Extracting functional keywords from the scheme word feature set to obtain a keyword feature group;

[0015] Generate a function word feature set according to the subject word feature group and the keyword feature group;

[0016] Performing module function clustering on the function word feature set to obtain a standard function word feature set;

[0017] Function mapping is performed on the standard function word feature set to obtain a platform function set.

[0018] Optionally, the module function clustering of the function word feature set to obtain a standard function word feature set includes:

[0019] Performing word frequency weighting and attention weighting on the function word feature set to obtain a weighted function word feature set;

[0020] Performing Gaussian distribution center screening on the weighted function word feature set to obtain a primary center word feature group;

[0021] Performing neighborhood division on the primary core word feature group according to the weighted function word feature set to obtain a primary word feature group set;

[0022] Selecting primary word feature groups in the primary word feature group set one by one as target primary word feature groups, and using the primary core word features corresponding to the target primary word feature groups in the primary core word feature group as target primary core word features;

[0023] The weighted function word features in the weighted function word feature set are selected one by one as the target weighted function word features, and the weighted module distance between the target weighted function word feature and the target primary word feature group is calculated according to the target primary center word feature using the following weighted module distance algorithm:

[0024]

[0025] Wherein, L refers to the weighted module distance, max is the maximum value symbol, i is the feature number, N is the total number of features of the target primary word feature group, q i is the weight of the i-th weighted function word feature in the target primary word feature group, α i is the i-th weighted function word feature in the target primary word feature group, β is the target weighted function word feature, qβ is the weight of the target weighted function word feature β, · is the inner product symbol, || is the modulus symbol, γ is the target primary center word feature, q γ is the weight of the target primary central word feature;

[0026] The weighted module distances of all target weighted function word features in the weighted function word feature set are collected into a weighted module distance set, and the target primary word feature group is threshold-compared and updated according to the weighted module distance set to obtain a target secondary word feature group;

[0027] All target secondary word feature groups are collected into a secondary word feature group set, and the secondary word feature group set is iteratively updated to obtain a standard function word feature set.

[0028] Optionally, performing functional feature matching and development specifications on multiple teams according to the platform function set to obtain specification development information includes:

[0029] Performing word vector encoding and feature matrixing on the platform function set to obtain a function feature matrix set;

[0030] Performing attention encoding on the functional feature matrix set to obtain an attention functional feature set;

[0031] Performing residual connection and layer normalization operations on the attention function feature set to obtain a normalized function feature set;

[0032] Performing feedforward linear transformation and feedforward activation on the normalized functional feature set to obtain an activated functional feature set;

[0033] Performing residual connection and layer normalization operations on the normalized function feature set and the activated function feature set to obtain a coding function feature set;

[0034] Attention decoding is performed on the encoded functional feature set to obtain standard development information.

[0035] Optionally, performing multi-team module development and specification review on the platform function set according to the specification development information to obtain a specification code set includes:

[0036] Selecting platform functions in the platform function set one by one as target platform functions, assigning the target platform functions to a team, and obtaining a target development team;

[0037] Sending the specification development information and the target platform function to the target development team, and obtaining feedback code from the target development team regarding the specification development information and the target platform function;

[0038] Compiling the specification development information into static rules to obtain static specification rules;

[0039] Performing a standard review on the feedback code according to the static standard rules to obtain a vulnerability code group and a review log;

[0040] Sending the vulnerability code group and the review log to the target development team, and obtaining a repair code group from the target development team for the vulnerability code group and the review log;

[0041] The feedback code is repaired and updated according to the repair code group to obtain a standard code, and the standard codes of all target platform functions in the platform function set are collected into a standard code group.

[0042] Optionally, the isolating and deploying the standard code set and integrating the micro-applications to obtain a supply chain front end includes:

[0043] Functionally packaging the standard code set to obtain a function code package set;

[0044] Performing container matching on the function code package set to obtain a function container set;

[0045] Performing code deployment on the function code package set according to the function container set to obtain a function module set;

[0046] Initializing the main application of the functional module set to obtain the module main application;

[0047] The micro-application is registered and the application is loaded for the functional module set according to the module main application to obtain the supply chain front end.

[0048] Optionally, extracting behavior features and analyzing behavior on the user behavior log to obtain analyzed user behavior includes:

[0049] Performing log analysis on the user behavior log to obtain user behavior data;

[0050] Performing time series behavior matching and behavior feature encoding on the user behavior data to obtain a behavior feature sequence;

[0051] Performing differential timing gate activation on the behavior feature sequence to obtain differential timing features;

[0052] Performing multi-head attention encoding on the behavior feature sequence to obtain a behavior attention feature;

[0053] Performing cyclic time series convolution on the behavior feature sequence to obtain convolution time series features;

[0054] Performing full connection fusion on the differential time series feature, the behavioral attention feature and the convolutional time series feature to obtain a fused behavioral feature;

[0055] Performing forward feedback propagation on the fusion behavior feature to obtain an analysis behavior feature;

[0056] The analyzed behavior features are subjected to a denormalization operation and a behavior mapping operation to obtain analyzed user behaviors.

[0057] Optionally, the performing full-connection fusion on the differential time series feature, the behavior attention feature and the convolution time series feature to obtain the fused behavior feature includes:

[0058] The differential time series feature, the behavioral attention feature and the convolutional time series feature are fully connected and fused using the following fully connected algorithm to obtain a fused behavioral feature:

[0059] U t =tanh(W g g t +W a a t +W c c t +b)

[0060] Among them, U t refers to the fusion behavior characteristics of the t time step, t is the time step index, tanh is the activation function symbol, g t refers to the differential time series feature at time step t, W g refers to the weight matrix of the differential time series features, a t refers to the behavioral attention feature at time step t, W a refers to the weight matrix of the behavioral attention feature, c t refers to the convolutional time series feature at time step t, W c refers to the weight matrix of the convolutional time series features, and b refers to the preset fully connected bias vector.

[0061] Optionally, dynamically collaboratively loading the supply chain front end according to the analyzed user behavior includes:

[0062] Extracting module behaviors from the analyzed user behaviors to obtain a functional module sequence;

[0063] Extracting timestamps from the functional module sequence to obtain module switching timestamps;

[0064] Collaboratively annotate the functional module sequence according to the module switching timestamp to obtain current functional module data and switching functional module data;

[0065] Performing timed module unloading on the supply chain front end according to the current function module data;

[0066] The supply chain front end is preloaded with modules according to the switching function module data.

[0067] In order to solve the above problems, the present invention also provides a micro-front-end supply chain platform collaboration system based on multi-team development, the system comprising:

[0068] The development specification module is used to perform functional analysis on the supply chain solution to be developed, obtain the platform function set, match the functional characteristics of multiple teams and develop specifications based on the platform function set, and obtain standardized development information;

[0069] A code review module, used to perform multi-team module development and specification review on the platform function set according to the specification development information to obtain a specification code set;

[0070] An isolated deployment module, used for isolating and deploying the standard code set and integrating micro-applications to obtain a supply chain front end;

[0071] A behavior analysis module is used to obtain user behavior logs using the supply chain front end, extract behavior features and analyze behavior on the user behavior logs, and obtain analyzed user behavior;

[0072] A collaborative loading module is used to dynamically collaboratively load the supply chain front end according to the analyzed user behavior.

[0073] The embodiments of the present invention can realize the module division and clarification of platform functions by performing functional analysis and development specifications, and can combine the self-attention method to efficiently analyze the corresponding development specifications, thereby improving development efficiency. By performing multi-team module development and specification review, it can realize independent development and corresponding development review of multiple teams under the same standard, thereby improving development efficiency. By performing isolated deployment and micro-application integration, it can ensure team independence and reduce coupling between teams. At the same time, modular independent deployment is realized, which improves the flexibility and scalability of maintenance of the supply chain front end and improves development efficiency.

[0074] By extracting behavioral features and analyzing behaviors, it is possible to analyze and predict the usage of various functional modules in the supply chain front end by users in future time periods, thereby improving the efficiency of collaborative loading. By performing dynamic collaborative loading, it is possible to achieve dynamic collaboration of micro-front-end applications and improve the efficiency of micro-front-end applications. Therefore, the collaborative method and system of the micro-front-end supply chain platform based on multi-team development proposed in the present invention can solve the problem of low efficiency in micro-front-end development collaboration. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 A schematic diagram of a process flow of a micro-front-end supply chain platform collaboration method based on multi-team development provided by an embodiment of the present invention;

[0076] Figure 2 A schematic diagram of a flow chart of a functional analysis process provided by an embodiment of the present invention;

[0077] Figure 3 A schematic diagram of a process for performing isolated deployment and micro-application integration provided by an embodiment of the present invention;

[0078] Figure 4 A functional module diagram of a micro-frontend supply chain platform collaboration system based on multi-team development provided by an embodiment of the present invention;

[0079] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0080] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0081] The embodiment of the present application provides a micro-front-end supply chain platform collaboration method based on multi-team development. The execution subject of the micro-front-end supply chain platform collaboration method based on multi-team development includes but is not limited to at least one of the electronic devices such as the server, terminal, etc. that can be configured to execute the method provided in the embodiment of the present application. In other words, the micro-front-end supply chain platform collaboration method based on multi-team development can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc. The server can be an independent server, or it can be a cloud server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and basic cloud computing services such as big data and artificial intelligence platforms.

[0082] Reference Figure 1 As shown, it is a flow chart of a micro-frontend supply chain platform collaboration method based on multi-team development provided by an embodiment of the present invention. In this embodiment, the micro-frontend supply chain platform collaboration method based on multi-team development includes:

[0083] S1. Perform functional analysis on the supply chain solution to be developed to obtain a platform function set, and perform functional feature matching and development specifications on multiple teams based on the platform function set to obtain standardized development information.

[0084] In detail, the supply chain solution refers to text information of a series of tools, systems and strategies designed to manage and optimize various links of the financial supply chain. The financial supply chain involves the entire process from raw material procurement, production and manufacturing, logistics and transportation to the delivery of final products to consumers. The supply chain solution can be collected and organized by supply chain personnel after brainstorming.

[0085] Specifically, the various platform functions in the platform function set refer to the functions that need to be implemented by the supply chain finance platform to be developed, including supply chain financing, payment settlement, supply chain management, data analysis and risk management, etc. The platform functions correspond to the various micro-applications in the micro-front end.

[0086] In detail, the specification development information includes the technical specifications and UI specifications that need to be followed when developing each micro-frontend of the supply chain platform. The technical specifications include code style, naming rules, directory structure, technology stack (such as React, Vue, Angular, etc.) and API interface specifications. The UI specifications include design systems, component libraries and style guides to ensure that the modules developed by each team are consistent in appearance and interaction.

[0087] In the embodiment of the present invention, refer to Figure 2 As shown, the supply chain solution to be developed is subjected to functional analysis to obtain a platform function set, including:

[0088] S21, perform text denoising, word segmentation and word vector encoding on the supply chain solution to be developed to obtain a solution word feature set;

[0089] S22, performing topic modeling on the solution word feature set to obtain a topic word feature group;

[0090] S23, extracting functional keywords from the scheme word feature set to obtain a keyword feature group;

[0091] S24, generating a function word feature set according to the subject word feature group and the keyword feature group;

[0092] S25, performing module function clustering on the function word feature set to obtain a standard function word feature set;

[0093] S26: Perform function mapping on the standard function word feature set to obtain a platform function set.

[0094] In detail, the text denoising and word segmentation refers to removing irrelevant texts in the supply chain plan and performing text segmentation on the supply chain plan from which the irrelevant texts have been removed. The text denoising method may be a keyword matching denoising method, and the text word segmentation method may be a word segmentation tool such as jieba.

[0095] Specifically, the word vector encoding method can be a bag-of-words model method, a TF-IDF method, or a model encoding method such as Word2Vec, GloVe, or BERT, and a topic model (Latent Dirichlet Allocation, LDA for short) can be used for topic modeling, which refers to extracting topics from text, where each topic represents a group of related functions.

[0096] In detail, the module function clustering is performed on the function word feature set to obtain a standard function word feature set, including:

[0097] Performing word frequency weighting and attention weighting on the function word feature set to obtain a weighted function word feature set;

[0098] Performing Gaussian distribution center screening on the weighted function word feature set to obtain a primary center word feature group;

[0099] Performing neighborhood division on the primary core word feature group according to the weighted function word feature set to obtain a primary word feature group set;

[0100] Selecting primary word feature groups in the primary word feature group set one by one as target primary word feature groups, and using the primary core word features corresponding to the target primary word feature groups in the primary core word feature group as target primary core word features;

[0101] The weighted function word features in the weighted function word feature set are selected one by one as the target weighted function word features, and the weighted module distance between the target weighted function word feature and the target primary word feature group is calculated according to the target primary center word feature using the following weighted module distance algorithm:

[0102]

[0103] Wherein, L refers to the weighted module distance, max is the maximum value symbol, i is the feature number, N is the total number of features of the target primary word feature group, q i is the weight of the i-th weighted function word feature in the target primary word feature group, α i is the i-th weighted function word feature in the target primary word feature group, β is the target weighted function word feature, q β is the weight of the target weighted function word feature β, · is the inner product symbol, || is the modulus symbol, γ is the target primary center word feature, q γ is the weight of the target primary central word feature;

[0104] The weighted module distances of all target weighted function word features in the weighted function word feature set are collected into a weighted module distance set, and the target primary word feature group is threshold-compared and updated according to the weighted module distance set to obtain a target secondary word feature group;

[0105] All target secondary word feature groups are collected into a secondary word feature group set, and the secondary word feature group set is iteratively updated to obtain a standard function word feature set.

[0106] In detail, the word frequency weighting method may be TF-IDF, TextRank and other methods, the attention weighting method may be a position encoding algorithm of a self-attention mechanism, and the Gaussian distribution center screening refers to calculating the Gaussian distribution of the weighted function word feature set, and screening out a number of non-adjacent weighted function word features from the weighted function word feature set according to the Gaussian distribution as primary center word features to form a primary center word feature group.

[0107] Specifically, the neighborhood division means that the Gaussian distribution of the weighted function word feature set selects a feature neighborhood for each primary central word feature group in the primary central word feature group, and collects the features in the feature neighborhood as primary word feature groups into a primary word feature group set.

[0108] In detail, the weighted module distance algorithm can combine the weights of word features and the mean distances between the target weighted functional word features and the target primary word feature group to determine whether the target price full-function word features meet the update requirements. The threshold comparison update and the iterative update method can be a density clustering algorithm (Ordering Points To Identify the Clustering Structure, abbreviated as OPTICS).

[0109] Specifically, the functional keyword extraction refers to extracting the functionally related feature words in the scheme word feature set to integrate the functional word feature set, wherein the functional keyword extraction method may be TF-IDF, TextRank and other methods.

[0110] In detail, the generating of the function word feature set refers to removing duplicates and merging the subject word feature group and the keyword feature group, and the function mapping method refers to mapping each standard function word feature class in the standard function word feature class set into a platform function using the reverse mapping scheme of the word variable encoding, and aggregating all the platform functions into a platform function set.

[0111] Specifically, the functional feature matching and development specifications of multiple teams are performed according to the platform function set to obtain standardized development information, including:

[0112] Performing word vector encoding and feature matrixing on the platform function set to obtain a function feature matrix set;

[0113] Performing attention encoding on the functional feature matrix set to obtain an attention functional feature set;

[0114] Performing residual connection and layer normalization operations on the attention function feature set to obtain a normalized function feature set;

[0115] Performing feedforward linear transformation and feedforward activation on the normalized functional feature set to obtain an activated functional feature set;

[0116] Performing residual connection and layer normalization operations on the normalized function feature set and the activated function feature set to obtain a coding function feature set;

[0117] Attention decoding is performed on the encoded functional feature set to obtain standard development information.

[0118] In detail, the attention encoding method can be an encoding method based on the Transformer or BERT model, the residual connection is a connection method that directly skips a sublayer and adds the input and output of the sublayer to alleviate the problem of gradient disappearance during model training, and the layer normalization refers to normalizing the data.

[0119] Specifically, the feedforward linear transformation refers to linear transformation using two linear transformation functions of a feedforward neural network (FFN for short), the feedforward activation refers to activation using an activation function of the feedforward neural network, the attention decoding method can be decoding using a decoder of a trained Transformer model, and the Transformer model can be trained using multiple standard development information sets determined by personnel and corresponding platform function sets.

[0120] In the embodiment of the present invention, by performing functional analysis and development specifications, the module division and clarification of platform functions can be achieved, and the corresponding development specifications can be efficiently analyzed in combination with the self-attention method, thereby improving development efficiency.

[0121] S2. Perform multi-team module development and specification review on the platform function set according to the specification development information to obtain a specification code set.

[0122] In detail, each specification code in the specification code set corresponds to the development code of a platform function in the platform function set under the specification development information, and the platform function corresponds to each micro-application in the micro-frontend.

[0123] In the embodiment of the present invention, the multi-team module development and specification review of the platform function set are performed according to the specification development information to obtain a specification code set, including:

[0124] Selecting platform functions in the platform function set one by one as target platform functions, assigning the target platform functions to a team, and obtaining a target development team;

[0125] Sending the specification development information and the target platform function to the target development team, and obtaining feedback code from the target development team regarding the specification development information and the target platform function;

[0126] Compiling the specification development information into static rules to obtain static specification rules;

[0127] Performing a standard review on the feedback code according to the static standard rules to obtain a vulnerability code group and a review log;

[0128] Sending the vulnerability code group and the review log to the target development team, and obtaining a repair code group from the target development team for the vulnerability code group and the review log;

[0129] The feedback code is repaired and updated according to the repair code group to obtain a standard code, and the standard codes of all target platform functions in the platform function set are collected into a standard code group.

[0130] In detail, the team allocation refers to selecting a development team that develops the target platform function from a number of independent development teams as a target development team, and the team allocation can be performed according to the historical development records of each development team.

[0131] Specifically, the feedback code refers to the functional code developed by the target development team for the target platform function under the guidance of the specification development information. The static rule compilation method can be a rule compilation based on methods such as ESLint or Stylelint, and specification review can be performed using methods such as Prettier, ESLint or Stylelint.

[0132] In the embodiment of the present invention, by carrying out multi-team module development and specification review, independent development and corresponding development review of multiple teams under the same standard can be achieved, thereby improving development efficiency.

[0133] S3. Isolate and deploy the standard code set and integrate micro-applications to obtain a supply chain front end.

[0134] In detail, the supply chain front end is a micro-application front end that realizes the needs of the supply chain financial platform. The micro-application front end (Micro-Frontend) is an architectural model that draws on the idea of ​​the microservice architecture and decomposes the front-end application into several independent and autonomous micro-applications (Micro-Applications). These micro-applications can be independently developed, tested, deployed and maintained by different teams, and then integrated through a main application to form an overall user interface.

[0135] In the embodiment of the present invention, refer to Figure 3 As shown, the supply chain front end is obtained by matching the standard code set to the standard code set, including:

[0136] S31, functionally packaging the standard code set to obtain a functional code package set;

[0137] S32, performing container matching on the function code package set to obtain a function container set;

[0138] S33, performing code deployment on the function code package set according to the function container set to obtain a function module set;

[0139] S34, initializing the main application of the functional module set to obtain the module main application;

[0140] S35. Register the micro-application and load the application for the functional module set according to the module main application to obtain the supply chain front end.

[0141] Specifically, the function packaging refers to packaging the standard code set and the corresponding resource code, and Webpack can be used for function packaging. The container matching refers to screening out the corresponding deployment container or server for each function code package in the function code package set. The code deployment method is to specify the corresponding function container in the function container set for each function code package in the function code package set.

[0142] In detail, each functional module in the functional module set is a micro-application module in the micro-front end, and the main application initialization includes initializing the routing address of the module main application, and generating a corresponding navigation configuration according to the routing address of each functional module in the functional module set.

[0143] Specifically, the module main application is responsible for managing and loading various micro-applications, while providing unified navigation and public resource micro-applications, and can use the Qiankun framework to register micro-applications and load applications.

[0144] In the embodiment of the present invention, by performing isolated deployment and micro-application integration, the independence of the team can be guaranteed and the coupling between teams can be reduced. At the same time, modular independent deployment is achieved, the flexibility and scalability of the maintenance of the supply chain front end are improved, and the development efficiency is improved.

[0145] S4. Utilize the supply chain front end to obtain user behavior logs, extract behavior features and perform behavior analysis on the user behavior logs, and obtain analyzed user behaviors.

[0146] In detail, the user behavior log refers to the tracking log obtained when the user uses the supply chain front-end. The user behavior log records information such as the timestamp, frequency, and duration of the user's use of each functional module in the supply chain front-end. The analyzed user behavior refers to the analyzed user's behavior of using the supply chain front-end in a future time period.

[0147] In the embodiment of the present invention, the extracting behavior features and analyzing the behavior of the user behavior log to obtain the analyzed user behavior includes:

[0148] Performing log analysis on the user behavior log to obtain user behavior data;

[0149] Performing time series behavior matching and behavior feature encoding on the user behavior data to obtain a behavior feature sequence;

[0150] Performing differential timing gate activation on the behavior feature sequence to obtain differential timing features;

[0151] Performing multi-head attention encoding on the behavior feature sequence to obtain a behavior attention feature;

[0152] Performing cyclic time series convolution on the behavior feature sequence to obtain convolution time series features;

[0153] Performing full connection fusion on the differential time series feature, the behavioral attention feature and the convolutional time series feature to obtain a fused behavioral feature;

[0154] Performing forward feedback propagation on the fusion behavior feature to obtain analysis behavior feature;

[0155] The analyzed behavior features are subjected to a denormalization operation and a behavior mapping operation to obtain analyzed user behaviors.

[0156] Specifically, regular expressions or log parsing libraries can be used to traverse log parsing. The time series behavior matching method refers to matching a behavior data sequence from the user behavior data in time sequence through regular expressions or keyword matching methods. The behavior feature encoding refers to feature encoding of the behavior data sequence. Each behavior feature in the behavior feature sequence corresponds to the behavior frequency, type of functional module used, and usage duration of each behavior data in the behavior data sequence.

[0157] In detail, the differential temporal gate activation refers to extracting features from the behavioral feature sequence through a gated recurrent unit network in a time domain differential manner, the multi-head attention encoding refers to feature encoding using a multi-head attention mechanism, and the recurrent temporal convolution refers to feature extraction using a recurrent temporal neural network model.

[0158] In detail, the differential time series feature, the behavioral attention feature and the convolutional time series feature are fully connected and fused to obtain the fused behavioral feature, including:

[0159] The differential time series feature, the behavioral attention feature and the convolutional time series feature are fully connected and fused using the following fully connected algorithm to obtain a fused behavioral feature:

[0160] U t =tanh(W g g t +W a a t +W c c t +b)

[0161] Among them, U t refers to the fusion behavior characteristics of the t time step, t is the time step index, tanh is the activation function symbol, g t refers to the differential time series feature at time step t, W g refers to the weight matrix of the differential time series features, a t refers to the behavioral attention feature at time step t, W a refers to the weight matrix of the behavioral attention feature, c t refers to the convolutional time series feature at time step t, W c refers to the weight matrix of the convolutional time series features, and b refers to the preset fully connected bias vector.

[0162] In detail, by utilizing the fully connected algorithm, it is possible to achieve the fusion of multiple time series features and feature standardization operations, thereby improving the accuracy of behavior analysis. The behavior analysis model pre-trained by a large number of fused behavior features and corresponding behavior features can be used for forward feedback propagation. The behavior mapping operation is the inverse operation of the behavior feature encoding.

[0163] In the embodiment of the present invention, by performing behavior feature extraction and behavior analysis, it is possible to analyze and predict the usage of each functional module in the supply chain front end by users in future time periods, thereby improving the efficiency of collaborative loading.

[0164] S5. Dynamically and collaboratively load the supply chain front end according to the analyzed user behavior.

[0165] In an embodiment of the present invention, the dynamically collaborative loading of the supply chain front end according to the analyzed user behavior includes:

[0166] Extracting module behaviors from the analyzed user behaviors to obtain a functional module sequence;

[0167] Extracting timestamps from the functional module sequence to obtain module switching timestamps;

[0168] Collaboratively annotate the functional module sequence according to the module switching timestamp to obtain current functional module data and switching functional module data;

[0169] Performing timed module unloading on the supply chain front end according to the current function module data;

[0170] The supply chain front end is preloaded with modules according to the switching function module data.

[0171] In detail, the module behavior extraction refers to extracting the usage time data for each functional module in the analyzed user behavior, and the module switching timestamp includes the start time and deactivation time of each functional module in the functional module sequence.

[0172] Specifically, the collaborative labeling refers to selecting the functional module in use and the corresponding deactivation time in the module switching timestamp as the current functional module data, the functional module that may be used in the future and the corresponding start time as the switching functional module data according to the current time. The unmount() function can be used to perform timed module unloading, and the mount() function can be used to preload the module.

[0173] In the embodiment of the present invention, dynamic collaborative loading can be performed to achieve dynamic collaboration of micro-frontend applications and improve the utilization efficiency of micro-frontend applications.

[0174] The embodiments of the present invention can realize the module division and clarification of platform functions by performing functional analysis and development specifications, and can combine the self-attention method to efficiently analyze the corresponding development specifications, thereby improving development efficiency. By performing multi-team module development and specification review, it can realize independent development and corresponding development review of multiple teams under the same standard, thereby improving development efficiency. By performing isolated deployment and micro-application integration, it can ensure team independence and reduce coupling between teams. At the same time, modular independent deployment is realized, which improves the flexibility and scalability of maintenance of the supply chain front end and improves development efficiency.

[0175] By extracting behavioral features and analyzing behaviors, it is possible to analyze and predict the usage of various functional modules in the supply chain front end by users in future time periods, thereby improving the efficiency of collaborative loading. By performing dynamic collaborative loading, dynamic collaboration of micro-front-end applications can be achieved, thereby improving the efficiency of micro-front-end applications. Therefore, the micro-front-end supply chain platform collaboration method based on multi-team development proposed in the present invention can solve the problem of low efficiency in micro-front-end development collaboration.

[0176] like Figure 4 As shown, it is a functional module diagram of a micro-front-end supply chain platform collaborative system based on multi-team development provided by an embodiment of the present invention.

[0177] The micro-frontend supply chain platform collaborative system 100 based on multi-team development described in the present invention can be installed in an electronic device. According to the functions implemented, the micro-frontend supply chain platform collaborative system 100 based on multi-team development can include a development specification module 101, a code review module 102, an isolation deployment module 103, a behavior analysis module 104 and a collaborative loading module 105. The module described in the present invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.

[0178] In this embodiment, the functions of each module / unit are as follows:

[0179] The development specification module 101 is used to perform functional analysis on the supply chain solution to be developed, obtain a platform function set, match functional features and development specifications for multiple teams according to the platform function set, and obtain standardized development information;

[0180] The code review module 102 is used to perform multi-team module development and specification review on the platform function set according to the specification development information to obtain a specification code set;

[0181] The isolation deployment module 103 is used to isolate and deploy the standard code set and integrate micro-applications to obtain a supply chain front end;

[0182] The behavior analysis module 104 is used to obtain user behavior logs using the supply chain front end, extract behavior features and analyze behavior on the user behavior logs, and obtain analyzed user behaviors;

[0183] The collaborative loading module 105 is used to dynamically collaboratively load the supply chain front end according to the analyzed user behavior.

[0184] In detail, each module described in the micro-front-end supply chain platform collaborative system 100 based on multi-team development in the embodiment of the present invention is used in the same manner as described above. Figures 1 to 3 The same technical means are used as the collaborative method of the micro-frontend supply chain platform developed by multiple teams, and can produce the same technical effects, so I will not go into details here.

[0185] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, systems and methods can be implemented in other ways. For example, the system embodiments described above are only schematic, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.

[0186] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0187] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0188] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0189] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0190] In addition, it is obvious that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. The multiple units or systems stated in the system embodiment can also be implemented by one unit or system through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any specific order.

[0191] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A micro-frontend supply chain platform collaboration method based on multi-team development, characterized in that: The method comprises: Perform functional analysis on the supply chain solution to be developed to obtain the platform function set, and perform functional feature matching and development specifications on multiple teams based on the platform function set to obtain standardized development information; Perform multi-team module development and specification review on the platform function set according to the specification development information to obtain a specification code set; The standard code set is deployed in isolation and micro-applications are integrated to obtain a supply chain front end; Utilizing the supply chain front end to obtain user behavior logs, extracting behavior features and analyzing behavior on the user behavior logs, and obtaining analyzed user behaviors; The supply chain front end is dynamically and collaboratively loaded based on the analyzed user behavior.

2. The micro-frontend supply chain platform collaboration method based on multi-team development as claimed in claim 1, characterized in that: The functional analysis of the supply chain solution to be developed is performed to obtain a platform function set, including: Perform text denoising, word segmentation and word vector encoding on the supply chain solution to be developed to obtain the solution word feature set; Performing topic modeling on the scheme word feature set to obtain a topic word feature group; Extracting functional keywords from the scheme word feature set to obtain a keyword feature group; Generate a function word feature set according to the subject word feature group and the keyword feature group; Performing module function clustering on the function word feature set to obtain a standard function word feature set; Function mapping is performed on the standard function word feature set to obtain a platform function set.

3. The micro-frontend supply chain platform collaboration method based on multi-team development as claimed in claim 2, characterized in that: The module function clustering of the function word feature set to obtain a standard function word feature set includes: Performing frequency weighting and attention weighting on the function word feature set to obtain a weighted function word feature set; Performing Gaussian distribution center screening on the weighted function word feature set to obtain a primary center word feature group; Performing neighborhood division on the primary core word feature group according to the weighted function word feature set to obtain a primary word feature group set; Selecting primary word feature groups in the primary word feature group set one by one as target primary word feature groups, and using the primary core word features corresponding to the target primary word feature groups in the primary core word feature group as target primary core word features; The weighted function word features in the weighted function word feature set are selected one by one as the target weighted function word features, and the weighted module distance between the target weighted function word feature and the target primary word feature group is calculated according to the target primary center word feature using the following weighted module distance algorithm: Wherein, L refers to the weighted module distance, max is the maximum value symbol, i is the feature number, N is the total number of features of the target primary word feature group, q i is the weight of the i-th weighted function word feature in the target primary word feature group, α i is the i-th weighted function word feature in the target primary word feature group, β is the target weighted function word feature, q β is the weight of the target weighted function word feature β, · is the inner product symbol, || is the modulus symbol, γ is the target primary center word feature, q γ is the weight of the target primary central word feature; The weighted module distances of all target weighted function word features in the weighted function word feature set are collected into a weighted module distance set, and the target primary word feature group is threshold-compared and updated according to the weighted module distance set to obtain a target secondary word feature group; All target secondary word feature groups are collected into a secondary word feature group set, and the secondary word feature group set is iteratively updated to obtain a standard function word feature set.

4. The micro-frontend supply chain platform collaboration method based on multi-team development as claimed in claim 1, characterized in that: The function feature matching and development specifications of multiple teams are performed according to the platform function set to obtain standard development information, including: Performing word vector encoding and feature matrixing on the platform function set to obtain a function feature matrix set; Performing attention encoding on the functional feature matrix set to obtain an attention functional feature set; Performing residual connection and layer normalization operations on the attention function feature set to obtain a normalized function feature set; Performing feedforward linear transformation and feedforward activation on the normalized functional feature set to obtain an activated functional feature set; Performing residual connection and layer normalization operations on the normalized function feature set and the activated function feature set to obtain a coding function feature set; Attention decoding is performed on the encoded functional feature set to obtain standard development information.

5. The micro-frontend supply chain platform collaboration method based on multi-team development as claimed in claim 1, characterized in that: The multi-team module development and specification review of the platform function set are performed according to the specification development information to obtain a specification code set, including: Selecting platform functions in the platform function set one by one as target platform functions, assigning the target platform functions to a team, and obtaining a target development team; Sending the specification development information and the target platform function to the target development team, and obtaining feedback code from the target development team regarding the specification development information and the target platform function; Compiling the specification development information into static rules to obtain static specification rules; Performing a standard review on the feedback code according to the static standard rules to obtain a vulnerability code group and a review log; Sending the vulnerability code group and the review log to the target development team, and obtaining a repair code group from the target development team for the vulnerability code group and the review log; The feedback code is repaired and updated according to the repair code group to obtain a standard code, and the standard codes of all target platform functions in the platform function set are collected into a standard code group.

6. The micro-frontend supply chain platform collaboration method based on multi-team development as claimed in claim 1, characterized in that: The standard code set is isolated and deployed and micro-applications are integrated to obtain a supply chain front end, including: Functionally packaging the standard code set to obtain a function code package set; Performing container matching on the function code package set to obtain a function container set; Performing code deployment on the function code package set according to the function container set to obtain a function module set; Initializing the main application of the functional module set to obtain the module main application; The micro-application is registered and the application is loaded for the functional module set according to the module main application to obtain the supply chain front end.

7. The micro-frontend supply chain platform collaboration method based on multi-team development as claimed in claim 1, characterized in that: The extracting behavior features and analyzing the behavior logs to obtain the analyzed user behavior includes: Performing log analysis on the user behavior log to obtain user behavior data; Performing time series behavior matching and behavior feature encoding on the user behavior data to obtain a behavior feature sequence; Performing differential timing gate activation on the behavior feature sequence to obtain differential timing features; Performing multi-head attention encoding on the behavior feature sequence to obtain a behavior attention feature; Performing cyclic time series convolution on the behavior feature sequence to obtain convolution time series features; Performing full connection fusion on the differential time series feature, the behavioral attention feature and the convolutional time series feature to obtain a fused behavioral feature; Performing forward feedback propagation on the fusion behavior feature to obtain an analysis behavior feature; The analyzed behavior features are subjected to a denormalization operation and a behavior mapping operation to obtain analyzed user behaviors.

8. The micro-frontend supply chain platform collaboration method based on multi-team development as claimed in claim 7, characterized in that: The step of performing full connection fusion on the differential time series feature, the behavior attention feature and the convolution time series feature to obtain a fused behavior feature includes: The differential time series feature, the behavioral attention feature and the convolutional time series feature are fully connected and fused using the following fully connected algorithm to obtain a fused behavioral feature: U=tanh(Wggt+Waa+W.c+b) Among them, U t refers to the fusion behavior characteristics of the t time step, t is the time step index, tanh is the activation function symbol, g t refers to the differential time series feature at time step t, W g refers to the weight matrix of the differential time series features, a t refers to the behavioral attention feature at time step t, W a refers to the weight matrix of the behavioral attention feature, c t refers to the convolutional time series feature at time step t, W c refers to the weight matrix of the convolutional time series features, and b refers to the preset fully connected bias vector.

9. The micro-frontend supply chain platform collaboration method based on multi-team development as claimed in claim 1, characterized in that: The dynamically collaborative loading of the supply chain front end according to the analyzed user behavior includes: Extracting module behaviors from the analyzed user behaviors to obtain a functional module sequence; Extracting timestamps from the functional module sequence to obtain module switching timestamps; Collaboratively annotate the functional module sequence according to the module switching timestamp to obtain current functional module data and switching functional module data; Performing timed module unloading on the supply chain front end according to the current function module data; The supply chain front end is preloaded with modules according to the switching function module data.

10. A micro-front-end supply chain platform collaborative system based on multi-team development, characterized in that: The system comprises: The development specification module is used to perform functional analysis on the supply chain solution to be developed, obtain the platform function set, match the functional characteristics of multiple teams and develop specifications based on the platform function set, and obtain standardized development information; A code review module, used to perform multi-team module development and specification review on the platform function set according to the specification development information to obtain a specification code set; An isolated deployment module, used for isolating and deploying the standard code set and integrating micro-applications to obtain a supply chain front end; A behavior analysis module is used to obtain user behavior logs using the supply chain front end, extract behavior features and analyze behavior on the user behavior logs, and obtain analyzed user behavior; A collaborative loading module is used to dynamically collaboratively load the supply chain front end according to the analyzed user behavior.

Images