Financial service-based component recommendation method, system and equipment and storage medium

Abstract

The invention relates to the technical field of data processing, in particular to a financial business-based component recommendation method, system and device and a storage medium, and the method comprises the steps: obtaining an input demand, and carrying out the analysis to obtain a corresponding technical feature and a financial business context feature; matching in a preset component database based on the technical features to obtain a candidate component set; performing real-time compliance admission verification on the candidate component set to screen out components meeting safety and compliance standards; performing multi-target optimization sorting on the screened candidate component set based on the context features of the financial service and the matching degree of the candidate components and the input requirements in the technical features; and returning the sorted component list to the user and automatically generating a dependency configuration code corresponding to each component in the component list. According to the method, the service priority and the compliance risk can be considered at the same time on the premise of ensuring the technical feasibility, so that the recommendation result not only conforms to the security and compliance standard of the financial industry, but also can meet the actual demand of a specific service scene.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to a component recommendation method, system, device and storage medium based on financial business. BACKGROUND

[0002] In the current financial technology system, a financial business system is usually composed of a large number of interrelated functional modules, and its core business (such as payment settlement, identity authentication, risk control, clearing and reconciliation, etc.) depends on the coordinated operation of a large number of underlying software components, including serialization components, encryption components, communication components, log components, and rule engine components, etc. Since the real-time, stability and high reliability requirements of financial business are extremely high, the selection of any underlying component will directly affect the performance, security and sustainable maintenance capability of the system; at the same time, when building and evolving the business system, financial institutions often face strict regulatory requirements and audit constraints, so when selecting components, developers not only need to consider whether the technical functions match, but also need to comprehensively evaluate the license compliance, security vulnerability and adaptation degree of the component to the specific financial business scenario. Therefore, the selection of components has become a key link in the process of building a financial business system, and its correctness is directly related to the stability, compliance and reliability of the business and the subsequent operation and maintenance cost.

[0003] In the prior art, the component recommendation method is generally oriented to the optimal technical solution, and its core logic is to sort and filter based on the matching degree on the technical level. For example, the system usually calculates the technical similarity between the component and the developer's demand according to the technical function description, performance index, version information, etc. of the component, and generates a recommendation list according to the similarity. The judgment basis comes from the technical parameters themselves, so the recommendation mechanism tends to select components that are technically feasible and commonly used in projects. However, since the existing technology only considers the matching on the technical level, it cannot judge the importance of the component in the financial business and whether it meets the safety and compliance requirements. For example, a component may be technically feasible, but it may have risks or license compliance in payment, clearing and other key businesses. Therefore, the traditional component recommendation only focuses on the technical level and cannot meet the comprehensive requirements of safety, compliance and business importance in the financial industry, resulting in a disconnection between the recommended components and the financial scenario. SUMMARY

[0004] The present application provides a component recommendation method, system, device and storage medium based on financial business, which can consider the business priority and compliance risk while ensuring technical feasibility, so that the recommended results meet the safety and compliance standards of the financial industry and meet the actual needs of specific business scenarios. The present application provides the following technical solutions: In a first aspect, the present application provides a component recommendation method based on financial business, the method comprising: obtaining an input requirement of a developer and analyzing the input requirement to obtain technical features corresponding to the input requirement and financial business context features; matching the technical features corresponding to the input requirement in a preset component database to obtain a candidate component set; performing real-time compliance access verification on the candidate component set to filter out components meeting safety and compliance standards; performing multi-objective optimization sorting on the filtered candidate component set based on the financial business context features and the matching degree of the candidate components and the input requirement in the technical features; returning a sorted component list to a user and automatically generating dependency configuration code corresponding to each component in the component list.

[0005] In a specific implementable scheme, the obtaining of the input requirement of the developer and the analysis of the input requirement to obtain the technical features corresponding to the input requirement and the financial business context features include: adopting a BERT-based lightweight NLP model to perform semantic analysis on input text, performing word segmentation and context coding on the input requirement in the form of text, mapping each word to a vector representation, and capturing the context dependency relationship between the words; extracting the content describing the use, performance or interface of the component in the text as technical features through key term matching and semantic classification, the technical features being used to represent the specific requirements of the component in the technical level; mapping the requirement to financial business context features by identifying the financial business keywords involved in the text and combining a preset business scenario classification table, the financial business context features being used to represent the financial business scenario and related business attributes to which the input requirement belongs.

[0006] In a specific implementable scheme, the matching of the technical features corresponding to the input requirement in the preset component database to obtain the candidate component set includes: calculating the comprehensive matching degree of the technical features and the technical features of each component in the component database through an interactive compatibility matching function , the interactive compatibility matching function being as follows: ; wherein, is the basic matching degree of the i-th technical feature of the component and the corresponding technical feature, is a feature key factor, is the feature compatibility of feature and feature , is the feature compatibility of feature and feature , Interaction weights It is a nonlinear enhancement index. The total number of features; The components in the component database are sorted according to the calculated comprehensive matching degree, and the components with matching degrees higher than the preset threshold are selected to form a candidate component set.

[0007] In one specific implementation scheme, the real-time compliance access verification of the candidate component set to filter out components that meet security and compliance standards includes: The compliance access verification is based on a built-in financial-grade rule base and processed in real time through the Drools rule engine; Each attribute of the component is compared with the preset conditions in the rule base; In the rule chain, components are judged in order to determine whether they meet security and compliance requirements, and the admission status of the components is marked according to the matching results; After verification, all components that meet security and compliance standards are selected from the candidate component set.

[0008] In a specific implementation scheme, the multi-objective optimization ranking of the filtered candidate component set based on the financial business context features and the matching degree between the candidate components and the input requirements in terms of technical features includes: The financial business context characteristics of each component are decomposed into several quantitative indicators, including business weight factors. Compliance matching degree All indicators are normalized. For each component in the candidate component set, calculate the comprehensive ranking score. ; Based on the overall ranking score The filtered candidate component set is sorted in descending order to generate the final sorted component list.

[0009] In one specific implementation scheme, the formula for calculating the comprehensive ranking score is as follows: ; in, Candidate components The overall ranking score; Candidate components The degree of comprehensive matching between the technical characteristics of the input requirements and the technical characteristics of the input requirements. For the quantity of financial business, For the component in the first Weighting factors for each business dimension For the component in the first Compliance compliance across business dimensions It is a random fine-tuning factor; , , All are adjustment coefficients; This is a component risk correction factor.

[0010] In one specific feasible implementation, the component risk correction factor The calculation formula is as follows: ; in, The number of risk assessment indicators. For the first The weight of each risk indicator To make components In the The raw values ​​of each risk indicator are mapped to a function of standardized scores; For components Vulnerability rating; For components The update frequency; This is due to component dependency risks.

[0011] Secondly, this application provides a component recommendation system based on financial business, which adopts the following technical solution: A component recommendation system based on financial business includes: The requirement parsing module is used to obtain the input requirements from the developers and parse the input requirements to obtain the technical features and financial business context features corresponding to the input requirements. The component matching module is used to match the technical features corresponding to the input requirements in a preset component database to obtain a set of candidate components; The component filtering module is used to perform real-time compliance access verification on the candidate component set and filter out components that meet security and compliance standards. The component sorting module is used to perform multi-objective optimization sorting on the filtered candidate component set based on the financial business context features and the matching degree between the candidate components and the input requirements in terms of technical features. The component feedback module is used to return a sorted list of components to the user and automatically generate the dependency configuration code for each component in the list.

[0012] Thirdly, this application provides an electronic device, the device including a processor and a memory; the memory stores a program, the program being loaded and executed by the processor to implement a component recommendation method based on financial business as described in the first aspect.

[0013] Fourthly, this application provides a computer-readable storage medium storing a program that, when executed by a processor, is used to implement a component recommendation method based on financial business as described in the first aspect.

[0014] By structurally parsing the natural language input requirements of developers, this application extracts clear technical features and financial business context features. Multi-dimensional matching is then performed in a pre-defined component database to form a candidate component set, which undergoes real-time compliance verification to ensure that the selected components meet the security and compliance requirements of the financial industry. Furthermore, this application combines the technical matching degree of candidate components with financial business context features, generating a final recommendation list through multi-objective optimization ranking, and automatically generating the dependency configuration code for each component, ensuring that the recommendation results can be directly applied to actual projects. Through this technical solution, the traditional component recommendation that only focuses on technical feasibility is extended into a comprehensive recommendation mechanism that considers technical matching, the importance of financial business, and compliance requirements. This effectively solves the problem of the disconnect between traditional component recommendation and financial business scenarios, achieving a balance between ensuring technical feasibility and considering business priority and compliance risks. The recommendation results not only meet the security and compliance standards of the financial industry but also satisfy the actual needs of specific business scenarios.

[0015] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, the preferred embodiments of this application are described in detail below with reference to the accompanying drawings. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the component recommendation method based on financial business in an embodiment of this application.

[0017] Figure 2 This is a schematic diagram of the overall process of the component recommendation method based on financial business in the embodiments of this application.

[0018] Figure 3 This is a structural block diagram of a component recommendation system based on financial business in an embodiment of this application.

[0019] Figure 4 This is a block diagram of an electronic device based on component recommendations for financial transactions, as described in this application embodiment. Detailed Implementation

[0020] The specific embodiments of this application will be described in further detail below with reference to the accompanying drawings and examples. The following examples are used to illustrate this application, but are not intended to limit the scope of this application.

[0021] Optionally, this application uses the component recommendation method based on financial business provided in various embodiments as an example for description in electronic devices. The electronic device is a terminal or a server. The terminal can be a computer, tablet computer, etc. This embodiment does not limit the type of electronic device.

[0022] Reference Figure 1 This is a flowchart illustrating a component recommendation method based on financial transactions provided in one embodiment of this application. The method includes at least the following steps: Step S101: Obtain the developer's input requirements and parse them to obtain the technical features and financial business context features corresponding to the input requirements.

[0023] In step S101, the component usage requirements input by the developer in natural language are received, and the input is semantically parsed to generate technical features and financial business context features. The technical features represent the specific technical requirements of the component, including component type (e.g., serialization component, encryption component), functional attributes (e.g., encryption support, high throughput performance), performance metrics (e.g., high performance, low latency), and interface form. The financial business context features characterize the financial business scenario and related business attributes to which the input requirements belong, such as payment, clearing, and identity authentication, and also indicate whether critical security or compliance functions are involved.

[0024] Specifically, an existing lightweight NLP model based on BERT is used to perform semantic parsing of the input text. The NLP model first segments and encodes the textual input requirements, mapping each word to a vector representation and capturing the contextual dependencies between words. Then, through key term matching and semantic classification, the content describing the component's purpose, performance, or interface is extracted as technical features. Simultaneously, by identifying financial business keywords in the text and combining them with a pre-defined business scenario classification table, the requirements are mapped to financial business contextual features, forming a structured information set. It should be noted that the above key term matching, semantic classification, and business keyword identification operations can all be implemented using existing natural language processing technologies and models without modifying the basic model structure.

[0025] For example, if a developer inputs a requirement of "a high-performance JSON serialization component for payment encryption," the semantic parsing process in step S101 can yield the following structured feature information: In terms of technical features, this includes the component type being a JSON serialization component, the functional attribute being encryption support, and the performance metric being high performance; in terms of financial business context features, this includes the business scenario being payment, and the business attribute involving key encryption operations. The above parsing results transform the input natural language requirement into a clear set of technical and business scenario attributes, thus forming a structured feature representation that the system can use.

[0026] Step S102: Match the technical features corresponding to the input requirements in the preset component database to obtain a set of candidate components.

[0027] In step S102, the technical features corresponding to the input requirements parsed in step S101 are first obtained, and then matched against a pre-defined component database to identify a set of candidate components that meet the input requirements. Technical features include component type, functional attributes, performance indicators, and interface forms, used to characterize the component's capabilities and attributes at the technical level. During the matching process, not only is the similarity between individual features and the input requirements considered, but the interaction relationships between various technical features and their criticality in specific financial business scenarios are also dynamically analyzed, thereby achieving a multi-dimensional, non-linear, and combinatorially dependent matching logic.

[0028] Specifically, the designed interactive compatibility matching function calculates the comprehensive matching degree between the technical features and the technical features of each component in the component database. The interactive compatibility matching function is as follows: ; in, For components The The basic matching degree between each technical feature and its corresponding technical feature is obtained by calculating the semantic similarity between the component attributes and the input requirement features using word vectors or sentence vectors for textual features; and by calculating the normalized difference or percentage difference for numerical features. This is a key feature factor used to reflect the importance of each technical feature in the current financial business scenario. This value can be dynamically assigned through a preset business scenario weight table. For example, in a payment scenario, encryption functionality and performance indicators are highly critical, therefore... Larger. Features With features Feature compatibility, used to quantify the internal characteristics of a component. The first feature and the second The synergistic adaptability of feature combinations can be obtained by analyzing the co-occurrence frequency and successful usage of each feature combination in historical projects in the component database. Features With features The interaction weights are used to measure the features corresponding to the input requirements. With features The importance of combinations in specific business scenarios can be assigned based on the sensitivity of financial business to feature combinations. This is a non-linear enhancement index used to amplify the priority of high-matching components in the final ranking, thereby enhancing the discriminative power of candidate components. The total number of features.

[0029] It should be noted that the reason the above interactive compatibility matching formula adopts a structure combining product and weighted superposition is based on the complexity and multidimensional dependencies of component selection in financial business. Firstly, the formula, through each technical feature... Perform index weighting By incorporating the criticality of features into the matching score calculation, the matching contribution of key features in specific business scenarios is amplified, thereby avoiding the problem of overall component inapplicability due to mismatch of key functions. Compared with traditional methods that rely solely on weighted averages or simple similarity accumulation, this exponential weighting enhances the influence of high-matching key features in the comprehensive score, improving the discriminative power of the matching results. Secondly, the formula introduces feature combination compatibility. Interaction weights The superposition term is used to quantify the combinational dependencies and collaborative adaptability between various technical features. This design addresses the actual needs of financial businesses: certain feature combinations (such as encryption and high-performance serialization) have higher business sensitivity and technical adaptability requirements in payment or clearing scenarios. Traditional matching methods typically only consider the matching degree of individual features, failing to reflect the combined effect between features, and are prone to situations where individual technical points have high matching but the overall combination is unsuitable. By introducing an interaction term, the formula can dynamically adjust the contribution of each feature combination to the final matching degree, thereby ensuring the reliability of candidate components in terms of overall technical adaptability.

[0030] Furthermore, the formula aggregates each feature and its combination using a product form and introduces a nonlinear enhancement coefficient. This approach amplifies the scores of components with high overall matching scores while suppressing the influence of components with low matching scores or unreasonable combinations. This non-linear aggregation method better reflects the advantages of highly matched components than linear weighting methods, improving the overall quality and ranking accuracy of the candidate component set. Regarding parameter design... Dynamically assign values ​​based on business scenarios, giving key features scenario-sensitive weighting. and This approach collectively characterizes the synergy and practical reliability of feature combinations. This multi-dimensional, non-linear, and combinatorially dependent structure ensures that key technical features occupy reasonable weights in matching degree calculations, increasing the priority of highly critical components. Furthermore, by leveraging feature combination compatibility and interaction weights, it captures multi-dimensional dependencies within components, making the candidate component set superior to traditional matching methods in terms of overall adaptability and business scenario fit. Compared to conventional methods, this formula significantly improves matching accuracy, candidate component quality, and ranking reliability in complex financial business scenarios, thus more accurately supporting the component selection needs of financial businesses.

[0031] Finally, the components in the component database are sorted based on the calculated comprehensive matching degree, and components with a matching degree higher than a preset threshold are selected to form a candidate component set. This step allows for the quantitative matching of the technical features corresponding to the input requirements with the actual components in the component database, thereby obtaining the candidate component set.

[0032] Step S103: Perform real-time compliance access verification on the candidate component set and select components that meet security and compliance standards.

[0033] In step S103, the candidate component set obtained in step S102 is first acquired, and compliance access verification is performed on each component. The compliance access verification is based on a built-in financial-grade rule base and processed in real-time by the Drools rule engine. Specifically, each attribute of the component, such as license type, vulnerability score (CVSS_score), and usage, is compared with the preset conditions in the rule base. The rules are sequentially used to determine whether the component meets the security and compliance requirements, and the access status of the component is identified based on the matching results. After verification, all components that meet the security and compliance standards are selected from the candidate component set.

[0034] For example: Reject access when the component license is GPL-3.0 (IF license == 'GPL-3.0' THEN REJECT); Reject access when the component has a high-risk vulnerability and a CVSS score greater than or equal to 7.0 (IF CVSS_score>= 7.0 THEN REJECT); Mark the component as requiring manual review when its use involves encryption operations but the license is not Apache-2.0 (IF component.usage == 'encryption' AND license != 'Apache-2.0' THEN FLAG_FOR_REVIEW).

[0035] Through the above operations, the candidate component set can be filtered to include only components that meet the security and compliance requirements of financial business, thus forming a reliable basic set of components and providing compliance assurance for subsequent ranking and recommendation. The compliance status of each component is clearly marked, making it easy to use directly in multi-objective ranking and recommendation outputs.

[0036] Step S104: Perform multi-objective optimization and ranking on the filtered candidate component set based on the financial business context features and the matching degree between the candidate components and the input requirements in terms of technical features.

[0037] In step S104, the first step is to obtain the set of candidate components that meet the security and compliance standards and the parsed financial business context features. Simultaneously, the technical matching degree calculated for each candidate component is obtained. Subsequently, a multi-objective optimization ranking was performed on the candidate component set, combining technical matching degree with financial business context features to determine the overall priority of the components.

[0038] Specifically, the financial business context characteristics of each component are first decomposed into several quantitative indicators, including business weight factors. Compliance matching degree Business weighting factors are automatically assigned based on the importance of the business system, such as 1.0 for the payment system and 0.3 for the back-end management system; compliance matching degree is calculated through the pass rate or score of the verification rules. All indicators are normalized to ensure consistent dimensions, facilitating non-linear calculations. Subsequently, a comprehensive ranking score is calculated for each component in the candidate component set. A higher score indicates a higher overall priority for the component in terms of technological fit, importance to financial business, and compliance. The formula for calculating the overall ranking score is as follows: ; in, Candidate components The overall ranking score; Candidate components The degree of comprehensive matching between the technical characteristics of the input requirements and the technical characteristics of the input requirements. For the quantity of financial business, For the component in the first Weighting factors for each business dimension For the component in the first Compliance compliance across business dimensions To introduce random fine-tuning factors, non-linearity in the sorting is added; , , These are all adjustment coefficients used to control the non-linear weights of technology matching, business / compliance coupling, and risk correction; This is a component risk correction factor, used to adjust the impact of potential component risks on the ranking. The calculation formula is as follows: ; in, The number of risk assessment indicators. For the first The weight of each risk indicator To make components In the The raw values ​​of each risk indicator are mapped to a function of standardized scores; For components Vulnerability scoring is performed by obtaining component data from public vulnerability databases or internal vulnerability scanning tools. The system retrieves the latest security vulnerability information and calculates the vulnerability severity based on the CVSS scoring standard to generate a numerical score. If a component has multiple vulnerabilities, the highest score or a weighted average score can be used as the numerical score. ; For components The update frequency is determined by the statistics component. The version update history, including the version release interval and the most recent update time, generates standardized values ​​based on the frequency and timeliness of updates, reflecting the maintenance activity of the component. Analyze component dependency risks. The system analyzes the dependencies and their security and compliance status, including vulnerability information, license types, and update status of dependent components. A numerical dependency risk score is generated by calculating the cumulative impact of potential risks in the dependency chain.

[0039] Ultimately based on The selected candidate component set is sorted in descending order to generate the final sorted component list. This method ensures technical compatibility while fully reflecting business priorities, compliance requirements, and risk control, achieving multi-objective optimized sorting.

[0040] In step S104, the set of candidate components that meet the selected safety and compliance standards is first obtained, and then combined with the technical matching degree calculated in step S102. The system also comprehensively sorts each component based on its financial business context characteristics. The goal of this sorting operation is to ensure a high degree of alignment between the component's technical features and the input requirements, while also considering the component's importance and compliance requirements across different financial business dimensions, and taking into account the component's potential risks. The final ranking is adjusted to form a multi-dimensional, non-linearly coupled ranking mechanism. The ranking process first adjusts the technical matching degree of each component exponentially, ensuring that the matching degree of technical features has a fundamental advantage in the ranking. Simultaneously, on the business dimension, the weighting factors of each dimension and the compliance matching degree are combined in a product form, and a random fine-tuning factor is added. A non-linear sorting mechanism is introduced to avoid the linear monotonicity problem inherent in simple weighting. Subsequently, a power exponent is used. Amplify the overall impact of business and compliance characteristics, and then use the denominator A non-linear correction is applied to the potential risks of components, effectively suppressing the ranking scores of high-risk components. Compared with traditional ranking methods based on technology matching or simple weighting, this formula has significant technical advantages: on the one hand, the matching degree of technical features... To ensure that components meet input requirements in terms of functionality and performance, and to guarantee optimal technology, the non-linear coupling of financial business context features and risk correction allows the ranking results to simultaneously reflect business priorities, compliance requirements, and security risks. This avoids the limitations of conventional linear weighting methods that ignore dimensional interactions or fail to reflect risk adjustment. Furthermore, by using powers and products to non-linearly combine each dimension, the distinguishability and refinement of the ranking are improved, maximizing the comprehensive adaptability of components in terms of technology, business, and compliance dimensions. This results in a final candidate component ranking that is highly compatible with financial business scenarios.

[0041] Step S105: Return the sorted list of components to the user and automatically generate the dependency configuration code for each component in the list.

[0042] In step S105, the list of components from the comprehensive ranking results is provided to the developers. The list is arranged from highest to lowest comprehensive ranking score, allowing users to quickly view the most suitable components based on multiple dimensions such as technical suitability, financial business priority, compliance, and risk control. Simultaneously, for each candidate component, corresponding dependency configuration code is automatically generated based on its programming language and dependency management tool. This dependency configuration code is a standardized project file, such as a Maven pom.xml snippet in a Java project and a requirements.txt file in a Python project, containing the component name, version number, and necessary dependency declarations, for direct import into the developer's project.

[0043] Specifically, the process of generating dependency configuration code first parses the version information, module name, and dependencies between candidate components, and then organizes them into directly usable configuration text according to the dependency management specifications used by the target project. For example, in Maven pom.xml, this will be automatically written... <dependency>Labels and their <groupid> 、 <artifactid>and <version>information; in Python requirements.txt, the component name and version number are listed. The purpose of this is to enable developers to directly introduce recommended components in the project without manually searching or writing dependency declarations, thereby speeding up the development process, reducing the risk of errors, and ensuring that the version of the component used is consistent with the ranking recommendation results.

[0044] In summary, in combination with Figure 2 By structurally analyzing the developer's natural language input requirements, the application extracts explicit technical features and financial business context features, and then performs multi-dimensional matching in the preset component database to form a candidate component set, and implements real-time compliance access verification to ensure that the selected components meet the requirements of the financial industry in terms of safety and compliance. Further, the application combines the technical matching degree of the candidate components with the financial business context features, generates a final recommendation list through multi-objective optimization sorting, and automatically generates the dependency configuration code corresponding to each component to ensure that the recommendation results can be directly applied to actual projects. Through the above technical solutions, the traditional component recommendation that only focuses on technical feasibility is expanded to a comprehensive recommendation mechanism that takes into account technical matching, financial business importance, and compliance requirements, thereby effectively solving the problem of the disconnection between traditional component recommendation and financial business scenarios, and realizing the consideration of business priority and compliance risk while ensuring technical feasibility, so that the recommendation results meet the safety and compliance standards of the financial industry and meet the actual needs of specific business scenarios.

[0045] Figure 3 is a structural diagram of a component recommendation system based on financial business provided by an embodiment of the application, and the system at least includes the following modules: The requirement analysis module is configured to obtain the input requirements of the developer and analyze the input requirements to obtain the technical features and financial business context features corresponding to the input requirements. The component matching module is configured to match the technical features corresponding to the input requirements in the preset component database to obtain a candidate component set. The component screening module is configured to perform real-time compliance access verification on the candidate component set to screen out components that meet the safety and compliance standards. The component sorting module is configured to perform multi-objective optimization sorting on the screened candidate component set based on the financial business context features and the matching degree of the candidate components and the input requirements in the technical features. The component feedback module is configured to return the sorted component list to the user and automatically generate the dependency configuration code corresponding to each component in the component list.

[0046] For related details, refer to the above method embodiments.

[0047] Figure 4 is a block diagram of an electronic device provided by an embodiment of the present application. The device at least includes a processor 401 and a memory 402.

[0048] The processor 401 can include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 401 can be implemented in at least one of a hardware form of a DSP (Digital Signal Processing), an FPGA (Field-Programmable Gate Array), a PLA (Programmable Logic Array). The processor 401 can also include a main processor and a coprocessor. The main processor is a processor for processing data in an awake state, also known as a CPU (Central Processing Unit). The coprocessor is a low-power processor for processing data in a standby state. In some embodiments, the processor 401 can be integrated with a GPU (Graphics Processing Unit) that is responsible for rendering and drawing the content required to be displayed by the display screen. In some embodiments, the processor 401 can also include an AI (Artificial Intelligence) processor for processing machine learning-related computing operations.

[0049] The memory 402 can include one or more computer-readable storage media, which can be non-transitory. The memory 402 can also include a high-speed random access memory, and a non-volatile memory such as one or more disk storage devices, flash storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 402 is used to store at least one instruction for being executed by the processor 401 to implement the financial service-based component recommendation method provided by the method embodiment of the present application.

[0050] In some embodiments, the electronic device can also optionally include a peripheral device interface and at least one peripheral device. The processor 401, the memory 402, and the peripheral device interface can be connected through a bus or a signal line. Each peripheral device can be connected to the peripheral device interface through a bus, a signal line, or a circuit board. Illustratively, the peripheral device includes, but is not limited to, a radio frequency circuit, a touch display screen, an audio circuit, and a power supply, etc.

[0051] Of course, the electronic device can also include fewer or more components, which are not limited in the present embodiment.

[0052] Optionally, the present application further provides a computer readable storage medium, and the computer readable storage medium stores a program, and the program is loaded and executed by a processor to implement the component recommendation method based on financial services of the method embodiment.

[0053] Optionally, the present application further provides a computer product, and the computer product comprises a computer readable storage medium, and the computer readable storage medium stores a program, and the program is loaded and executed by a processor to implement the component recommendation method based on financial services of the method embodiment.

[0054] The technical features of the above embodiments can be combined in any manner. To make the description concise, all possible combinations of the technical features in the above embodiments are not described, but as long as the combinations of the technical features do not exist contradictions, they should be considered as the scope of the present disclosure.

[0055] The above embodiments only express several implementation manners of the present application, and the description is specific and detailed, but it should not be understood as a limitation on the scope of the patent. It should be pointed out that for those skilled in the art, without departing from the concept of the present application, some modifications and improvements can be made, which are all within the protection scope of the present application. Therefore, the protection scope of the patent of the present application should be subject to the appended claims.< / version> < / artifactid> < / groupid> < / dependency>

Claims

1. A method of recommending components based on financial services, characterized by, The method comprises: obtaining the input requirements of the developer and analyzing the input requirements to obtain technical features corresponding to the input requirements and financial business context features; matching the technical features corresponding to the input requirements in a preset component database to obtain a candidate component set; performing real-time compliance access verification on the candidate component set to screen out components meeting safety and compliance standards; performing multi-objective optimization sorting on the screened candidate component set based on the financial business context features and the matching degree of the candidate components and the input requirements in technical features; returning the sorted component list to the user and automatically generating the dependency configuration code corresponding to each component in the component list.

2. The component recommendation method based on financial services according to claim 1, wherein, The method comprises: performing semantic analysis on the input text by using a BERT-based lightweight NLP model, performing word segmentation and context coding on the input requirements in the form of text, mapping each word to a vector representation, and capturing the context dependency relationship between the words; extracting the content describing the use, performance or interface of the component in the text as technical features by key term matching and semantic classification, wherein the technical features are used to represent the specific requirements of the component in the technical layer; mapping the requirements to financial business context features by identifying the financial business keywords involved in the text and combining a preset business scenario classification table, wherein the financial business context features are used to represent the financial business scenarios and related business attributes to which the input requirements belong.

3. The financial service-based component recommendation method of claim 1, wherein, The method comprises: The comprehensive matching degree of the technical features and each component technical feature in the component database is calculated through an interactive compatible matching function The interactive compatible matching function is as follows: ； wherein, is a component of a first technical feature that is a basic match to the corresponding technical feature, is a feature criticality factor, is a feature compatibility of the feature , is an interaction weight of the feature with the feature , is a non-linear enhancement exponent, is a total number of features; sorting the components in the component database according to the calculated comprehensive matching degree, selecting the components with a matching degree higher than a preset threshold, and constructing the candidate component set.

4. The financial service-based component recommendation method of claim 1, wherein, The method comprises: The compliance access verification is based on an embedded financial-level rule library and is processed in real time by a Drools rule engine; each attribute of the component is compared with the preset conditions in the rule library one by one; whether the component meets the safety and compliance requirements is determined in the rule chain according to the order, and the access state of the component is identified according to the matching result; after the verification is completed, all components meeting the safety and compliance standards are screened out from the candidate component set.

5. The financial service-based component recommendation method of claim 3, wherein, The method comprises: The financial service context features of each component are decomposed into a plurality of quantitative indexes, including business weight factors and compliance matching degree , and all indexes are normalized. calculating an overall ranking score for each component of the candidate component set ; According to the comprehensive ranking score The screened candidate component set is arranged in descending order to generate a final ranked component list.

6. The financial service-based component recommendation method according to claim 5, wherein The calculation formula of the comprehensive sorting score is as follows: ； wherein, is a composite ranking score for the candidate component ; is a composite match score for the technical features of the candidate component to the input requirement technical features, is the number of financial service dimensions, is the weight factor of the component in the th service dimension, is the compliance match score of the component in the th service dimension, is a random fine-tuning factor; , , are all adjustment coefficients; is a component risk correction factor.

7. The financial service-based component recommendation method according to claim 6, wherein, The component risk modifier The formula for calculating the component risk modifier is as follows: ； wherein, is the number of risk assessment indicators, is the weight of the risk indicator, is a function that maps the raw value of the risk indicator to a normalized score; is the vulnerability score of the component, is the update frequency of the component, is the component dependency risk.​​​ 8. A component recommendation system based on financial services, characterized by, The method comprises: a requirement analysis module, configured to obtain the input requirements of the developer and analyze the input requirements to obtain technical features corresponding to the input requirements and financial business context features; a component matching module, configured to match the technical features corresponding to the input requirements in a preset component database to obtain a candidate component set; a component screening module, configured to perform real-time compliance access verification on the candidate component set to screen out components meeting safety and compliance standards; and a component sorting module, configured to perform multi-objective optimization sorting on the screened candidate component set based on the financial business context features and the matching degree of the candidate components and the input requirements in technical features. a component ranking module configured to perform multi-objective optimization ranking on the filtered candidate component set based on the financial business context features and the matching degree of the candidate components and the input requirements in technical features; a component feedback module configured to return the ranked component list to the user and automatically generate the dependent configuration code corresponding to each component in the component list.

9. An electronic device, comprising: The device comprises a processor and a memory; the memory stores a program, which is loaded and executed by the processor to implement the financial business-based component recommendation method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The storage medium stores a program, which is executed by the processor to implement the financial business-based component recommendation method according to any one of claims 1 to 7.

Images