Enterprise portrait generation method and device, electronic equipment and storage medium

Abstract

The invention discloses an enterprise portrait generation method and device, electronic equipment and a storage medium. Comprising the steps that input information is obtained, semantic analysis is conducted on the input information, semantic information is obtained, and the semantic information comprises an enterprise name and an intention type; retrieving in a pre-constructed enterprise knowledge base based on the enterprise name and the intention type to obtain a target retrieval result; determining a basic scoring result through a rule engine based on the target retrieval result; an expert scoring rule is obtained, a target enterprise portrait is generated through a pre-constructed enterprise portrait generation model based on the expert scoring rule, the target retrieval result and the basic scoring result, and the target enterprise portrait is generated according to a preset enterprise portrait template. The target enterprise portrait comprises the total score result, the score value of the at least one score index and score interpretation information corresponding to each score value. According to the scheme, the enterprise portrait containing detailed scores and explanations is generated by relying on knowledge base retrieval and combining the rule engine and the expert rule.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method, apparatus, electronic device, and storage medium for generating enterprise profiles. Background Technology

[0002] With the deepening of the digital economy, the information generated by enterprise operations is exploding. To determine accurate enterprise profiles, it is necessary to obtain relevant information from multiple dimensions or channels. Currently, existing enterprise profile generation methods cannot quickly obtain the comprehensive information needed to generate enterprise profiles, nor can they effectively process the acquired multi-dimensional heterogeneous data. The enterprise scoring models used cannot perform validity analysis on the scoring results and lack interpretability of the enterprise profiles. As a result, the obtained enterprise profiles lack effective scoring results and explanatory information, have poor usability, and cannot meet users' needs for enterprise profiles. Summary of the Invention

[0003] This invention provides a method, apparatus, electronic device, and storage medium for generating enterprise profiles, in order to solve the problems of being unable to obtain effective enterprise profiles and the poor interpretability of enterprise profiles.

[0004] According to one aspect of the present invention, a method for generating enterprise profiles is provided, comprising:

[0005] Obtain input information, perform semantic parsing on the input information to obtain semantic information, which includes the company name and intent type;

[0006] The search results are obtained by performing a search on a pre-built enterprise knowledge base based on the enterprise name and intent type.

[0007] The basic scoring results are determined using a rule engine based on the target search results;

[0008] The system obtains expert scoring rules, and generates a target enterprise profile based on the expert scoring rules, target search results, and basic scoring results using a pre-built enterprise profile generation model. The target enterprise profile is generated according to a preset enterprise profile template and includes a total score, a score value for at least one scoring indicator, and a score explanation for each score value.

[0009] Optionally, semantic parsing is performed on the input information to obtain semantic information, including: preprocessing the input information to obtain preprocessed data, wherein the preprocessing includes text cleaning, word segmentation, and stop word filtering; extracting the enterprise name and key fields from the preprocessed data; and determining the intent type corresponding to the input information based on the key fields through a pre-built intent recognition model, wherein the intent type is a type set based on demand information, including user attention dimensions, information depth, and application scenarios.

[0010] Optionally, a search is performed in a pre-built enterprise knowledge base based on the enterprise name and intent type to obtain the target search results, including: searching the business registration information in the enterprise knowledge base based on the enterprise name to determine the enterprise identifier and basic enterprise information corresponding to the enterprise name; calling the intent-to-dimensional mapping interface based on the intent type to obtain at least one dimension corresponding to the intent type; searching the enterprise knowledge base based on the enterprise identifier and at least one dimension to determine the data details corresponding to each dimension; and performing fusion processing based on the enterprise identifier, basic enterprise information, at least one dimension, and the data details corresponding to each dimension to obtain the target search results.

[0011] Optionally, determining the basic scoring result based on the target retrieval result using a rule engine includes: calling the rule engine and loading basic scoring rules, wherein the basic scoring rules include the correspondence between dimensions and scoring rules; inputting the target retrieval result into the rule engine, parsing the target retrieval result through the rule engine, and extracting data details for at least one dimension; matching at least one dimension with the basic scoring rules, and determining the scoring result for each dimension based on the successfully matched scoring rules and the data details for each dimension; and performing fusion processing based on the scoring results for each dimension to determine the basic scoring result, wherein the basic scoring result includes the score value of at least one basic scoring indicator and the scoring explanation information corresponding to the score value of each basic scoring indicator.

[0012] Optionally, a target enterprise profile is generated based on expert scoring rules, target retrieval results, and basic scoring results using a pre-built enterprise profile generation model. This includes: obtaining a pre-built enterprise profile generation prompt template; determining enterprise profile generation prompt information based on expert scoring rules, basic scoring results, and the enterprise profile generation prompt template; inputting the enterprise profile generation prompt information and target retrieval results into the pre-built enterprise profile generation model; and having the pre-built enterprise profile generation model output the target enterprise profile.

[0013] Optionally, the method further includes: verifying the basic scoring results based on a first preset verification rule; if the basic scoring results do not meet the first preset verification condition, invoking a rule engine optimization processing method to optimize the rule engine, wherein the first preset verification rule includes integrity verification rules and / or consistency verification rules; and / or verifying the target enterprise profile based on a second preset verification rule; if the target enterprise profile does not meet the second preset verification condition, invoking an enterprise profile generation model optimization processing method to optimize the pre-built enterprise profile generation model, wherein the second preset verification rule includes enterprise profile information integrity verification rules.

[0014] Optionally, the method also includes: obtaining timestamp information from multiple data sources corresponding to a pre-built enterprise knowledge base; if the timestamp information from any data source is not the latest publication time, obtaining the latest published content from the data source; and updating the pre-built enterprise knowledge base based on the latest published content.

[0015] According to another aspect of the present invention, an apparatus for generating corporate profiles is provided, comprising:

[0016] The semantic information determination module is used to acquire input information, perform semantic parsing on the input information, and obtain semantic information, which includes the company name and intent type.

[0017] The target retrieval result determination module is used to search a pre-built enterprise knowledge base based on the enterprise name and intent type to obtain the target retrieval results.

[0018] The basic scoring result determination module is used to determine the basic scoring result based on the target retrieval result through a rule engine.

[0019] The target enterprise profile generation module is used to obtain expert scoring rules, and generate a target enterprise profile based on the expert scoring rules, target search results and basic scoring results through a pre-built enterprise profile generation model. The target enterprise profile is generated according to a preset enterprise profile template and includes the total score, the score value of at least one scoring indicator and the scoring explanation information corresponding to each score value.

[0020] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0021] At least one processor; and

[0022] A memory that is communicatively connected to at least one processor; wherein,

[0023] The memory stores a computer program that can be executed by at least one processor, such that the at least one processor is able to execute the enterprise profile generation method of any embodiment of the present invention.

[0024] According to another aspect of the present invention, a computer-readable storage medium is provided, which stores computer instructions for causing a processor to execute and implement the enterprise profile generation method of any embodiment of the present invention.

[0025] The technical solution of this invention involves acquiring input information, performing semantic parsing on the input information to obtain semantic information, wherein the semantic information includes the company name and intent type; performing a search in a pre-built enterprise knowledge base based on the company name and intent type to obtain target search results; determining basic scoring results based on the target search results using a rule engine; acquiring expert scoring rules; and generating a target company profile based on the expert scoring rules, the target search results, and the basic scoring results using a pre-built company profile generation model, wherein the target company profile is generated according to a preset company profile template and includes a total score, a score value for at least one scoring indicator, and scoring explanation information corresponding to each score value. This solution accurately extracts company names and intent types through semantic parsing of input information, ensuring the relevance and accuracy of subsequent searches. Relying on a pre-built company knowledge base provides comprehensive and reliable basic data support for scoring and profile generation. The basic scoring results determined by the rule engine guarantee the objectivity and consistency of the scoring, while incorporating expert scoring rules supplements professional experience judgment, making the scoring both data rigor and domain depth. The company profile generation model generates target company profiles containing total scores, specific scores, and corresponding explanations according to preset templates. This achieves both structured and standardized profiles, and improves readability through clear scoring logic, solving the problems of not being able to obtain effective company profiles and poor interpretability. The entire process forms a closed loop from information input to result output, balancing automation efficiency and professional reliability, providing comprehensive, accurate, and easy-to-understand references for relevant corporate decisions.

[0026] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0027] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0028] Figure 1 This is a flowchart of a method for generating enterprise profiles provided in Embodiment 1 of the present invention;

[0029] Figure 2 This is a flowchart of a method for generating enterprise profiles provided in Embodiment 2 of the present invention;

[0030] Figure 3 This is a schematic diagram of the structure of an enterprise profile generation device provided in Embodiment 3 of the present invention;

[0031] Figure 4 This is a schematic diagram of the structure of an electronic device that implements the enterprise profile generation method of this invention. Detailed Implementation

[0032] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0033] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0034] Example 1

[0035] Figure 1 This is a flowchart of a method for generating a corporate profile according to Embodiment 1 of the present invention. This embodiment is applicable to situations involving the generation of corporate profiles. This method can be executed by a corporate profile generation device, which can be implemented in hardware and / or software. The corporate profile generation device can be configured in electronic devices such as computers and servers. Figure 1 As shown, the method includes:

[0036] S110. Obtain input information, perform semantic parsing on the input information to obtain semantic information, which includes the company name and intent type.

[0037] The input information can be understood as information entered through a pre-set input method. For example, it could be information entered by the user in a designated area of ​​the visual interface, or voice information recorded through a microphone. The input method is not limited here. Semantic information can be understood as the core content obtained after semantic parsing of the input information. Semantic information is a refined and structured presentation of the input information, accurately reflecting its true meaning and core intent. It can be obtained by semantic extraction from the input information using natural language processing methods. In this embodiment, semantic information includes, but is not limited to, the company name and intent type. The intent type specifically represents the specific needs or target direction of the company profile conveyed through the input information. It reflects what problems the user hopes to solve or what information they hope to obtain through the company profile, such as understanding the company's credit rating, operating status, industry competitiveness, or specific business cooperation potential. Intent types include, but are not limited to, credit assessment, business analysis, risk identification, cooperation potential, and comprehensive evaluation. It should be noted that the intent type can be set according to the actual needs of generating the company profile; it is not limited here.

[0038] Specifically, the system receives raw input information from users in the form of text, speech-to-text, etc., and then processes the raw input information using natural language processing (NLP) technology to extract key elements, ultimately obtaining semantic information containing the company name and intent type, thus achieving a structured transformation of the raw input information. For example, NLP technology includes word segmentation, entity recognition, and intent type recognition.

[0039] In this embodiment, semantic parsing can accurately capture the core content of the input information, filter out irrelevant and interfering information, and transform unstructured raw input into structured semantic information that the system can directly process. This lays a precise foundation for subsequent steps such as retrieval, scoring, and profile generation based on enterprise name and intent type. At the same time, it improves the efficiency and accuracy of the entire process in understanding user input, ensuring that the system can respond quickly and meet the user's real needs.

[0040] Optionally, semantic parsing is performed on the input information to obtain semantic information, including: preprocessing the input information to obtain preprocessed data, wherein the preprocessing includes text cleaning, word segmentation and stop word filtering; extracting the enterprise name and key fields from the preprocessed data; and determining the intent type corresponding to the input information based on the key fields through a pre-built intent recognition model.

[0041] The intent type is a set of parameters based on demand information, which includes, but is not limited to, user focus dimensions, information depth, and application scenarios. Specifically, the intent type represents the specific direction of the user's demand for the enterprise profile content, clearly indicating the specific dimensions of information the user hopes to obtain from the enterprise profile or the problems that need to be solved. A corresponding intent type can be pre-set based on one or more of the user's focus dimensions, information depth, and application scenarios to accurately define the core purpose behind the input information. For example, in a business scenario, it might include cooperation consultation, product procurement, and complaint suggestions; in an information query scenario, it might cover enterprise qualification inquiries and industry data acquisition. It is a structured classification of the intent carried by the user's input information, enabling the system to quickly understand the user's core needs and providing clear guidance for subsequent accurate responses.

[0042] Specifically, after receiving the input information, it undergoes preprocessing. Text cleaning removes redundant symbols, erroneous characters, and other interfering content. Then, word segmentation breaks the text down into individual words, followed by filtering out meaningless stop words to obtain standardized preprocessed data. Next, the company name and key fields reflecting user needs are accurately extracted from the preprocessed data. Finally, the key fields are input into a pre-built intent recognition model. This model analyzes the semantic relationships and features of the key fields to determine and output the intent type corresponding to the input information. Ultimately, the company name and intent type are integrated to form semantic information. The intent recognition model can be a machine learning-based classification model.

[0043] In this embodiment, the preprocessing stage lays a high-quality data foundation for subsequent processing. Extracting the company name and key fields focuses on core elements. The combination of intent type set based on demand information and pre-built model makes the determination of semantic information more in line with actual application scenarios. This not only ensures the accuracy of the results but also improves the matching degree between semantic information and users' real needs. This makes the system more targeted, efficient, and consistent in information processing, and can better serve decision-making and response in various business scenarios.

[0044] S120. Based on the company name and intent type, perform a search in the pre-built company knowledge base to obtain the target search results.

[0045] Specifically, the target search results can be understood as a collection of information highly matched to the needs of specific enterprises and users, obtained through targeted retrieval from a pre-built enterprise knowledge base. This is achieved by using the parsed enterprise name as a precise positioning identifier and the intent type as a directional filtering criterion. It meets the dual conditions of enterprise name and intent type, directly responding to the user's core needs and possessing strong targeting and high value. The enterprise knowledge base can be understood as a dynamic data knowledge base integrating multi-dimensional internal and external information. It gathers enterprise data from different sources and dimensions, including but not limited to basic information, operational data, credit information, industry qualifications, and related dynamic information. Standardized data formats and association rules enable the orderly organization and efficient retrieval of information. Basic information includes but is not limited to enterprise name, registration information, and legal representative information; operational data includes but is not limited to revenue, profit, employee size information, and bidding information; credit information includes but is not limited to records of dishonesty and credit ratings; industry qualifications include but are not limited to license information and certification information; and related dynamic information includes but is not limited to investment and merger information, litigation disputes, administrative penalties, and public opinion information. The enterprise knowledge base can be pre-built using knowledge base construction methods, so that it can be directly accessed when generating enterprise profiles. It should be noted that the enterprise knowledge base is a dynamic knowledge base; scheduled verification and update tasks can be set to update the enterprise knowledge base and ensure that the information in the knowledge base is up-to-date, accurate, and complete.

[0046] Specifically, the parsed company name serves as the unique search identifier, precisely locating the corresponding information entry for that company in the company knowledge base, eliminating interference from other companies. Then, based on the intent type, the specific dimensions and scope of the search are determined. Through the pre-set indexing mechanism and association rules in the knowledge base, highly relevant sub-data (such as credit ratings, financial statements, risk records, etc.) is filtered from the company's information entries. This achieves precise location and extraction of information that meets both the company name and intent type criteria. This data is then integrated and structured to ultimately generate the target search results.

[0047] In this embodiment, searching the enterprise knowledge base by enterprise name ensures precise focus of the search scope and avoids interference from irrelevant enterprise information. The intent type further narrows the search dimensions, so that the results closely match the user's core needs, greatly improving search efficiency and accuracy. At the same time, relying on the pre-built enterprise knowledge base, the systematicness and authority of the information can be guaranteed, allowing users to quickly obtain targeted and valuable content, effectively reducing information filtering costs.

[0048] In some embodiments, the construction of an enterprise knowledge base can be achieved through a basic data processing module, a vectorization processing module, and a storage module. The basic data processing module performs data preprocessing on external data, with the following specific steps: Data cleaning: removing duplicate, invalid, and outdated data. External data includes, but is not limited to, company information, legal information, bidding information, public opinion information, company relationships, and company change information. Data segmentation: dividing long documents into smaller paragraphs or sentences to facilitate subsequent retrieval and generation. Based on the unified social credit code and company names, data silos are resolved, and cross-data source alignment is achieved. The vectorization processing module converts text data into vector representations (using Sentence-Transformer), and the storage module stores the text data and corresponding vectors in an Elasticsearch database to obtain the enterprise knowledge base. It should be noted that a hybrid search approach can be used when searching the enterprise knowledge base, that is, combining Elasticsearch's keyword and vector search to obtain the target search results. This search method can improve the search accuracy. For example, keyword matching (such as "company name + risk") and vector similarity search (encoding the company name and / or intent type into a vector and matching similar company event descriptions).

[0049] Optionally, a search is performed in a pre-built enterprise knowledge base based on the enterprise name and intent type to obtain the target search results, including: searching the business registration information in the enterprise knowledge base based on the enterprise name to determine the enterprise identifier and basic enterprise information corresponding to the enterprise name; calling the intent-to-dimensional mapping interface based on the intent type to obtain at least one dimension corresponding to the intent type; searching the enterprise knowledge base based on the enterprise identifier and at least one dimension to determine the data details corresponding to each dimension; and performing fusion processing based on the enterprise identifier, basic enterprise information, at least one dimension, and the data details corresponding to each dimension to obtain the target search results.

[0050] Specifically, a search can be conducted in the business registration information of the enterprise knowledge base based on the enterprise name to accurately locate and determine the corresponding enterprise identifier (such as the unified social credit code) and basic enterprise information. Then, based on the intent type, the intent-to-dimensional mapping interface is called to obtain at least one dimension associated with that intent type (such as the credit assessment intent corresponding to dimensions such as "dishonesty record" and "credit rating"). Subsequently, using the enterprise identifier as the unique subject identifier, a search is conducted in the knowledge base in combination with the above dimensions. The search method can adopt RAG (Retrieval-Augmented Generation) retrieval or BM25 retrieval algorithm to obtain specific data details under each dimension (such as the number of defaults and time under the "dishonesty record" dimension). Finally, the enterprise identifier, basic enterprise information, each dimension and its corresponding data details are integrated and merged to form a complete and structured target search result.

[0051] In this embodiment, through step-by-step precise retrieval and mapping, the uniqueness and accuracy of information are ensured by relying on enterprise identification, and the search scope is precisely focused by leveraging the association between intent type and dimension, avoiding redundant output of irrelevant information. At the same time, the fusion processing of multi-dimensional data ensures that the target search results include both basic information and in-depth content, which can comprehensively respond to user needs and greatly improve the relevance, completeness and practical value of the search results, especially in complex business scenarios where it can efficiently support user decision-making.

[0052] S130. Determine the basic scoring results based on the target retrieval results using a rule engine.

[0053] Specifically, a rule engine can be understood as an automated processing system that follows a set of pre-defined and executable rules. These executable rules are formulated based on business logic, evaluation criteria, and other information. They receive input data and perform logical judgments, calculations, or decisions according to the rules, achieving automated processing of specific tasks. They are also flexible, configurable, logically transparent, and easy to maintain, and can quickly respond to adjustments in business rules. In this embodiment, the rule engine can set basic scoring rules to process target retrieval results and determine basic scoring results. The basic scoring results can be understood as an initial quantitative evaluation derived by the rule engine based on the target retrieval results and according to the pre-defined basic scoring rules. It represents a basic quantitative presentation of the enterprise in a specific evaluation scenario (such as credit, operational capabilities, etc.), providing standardized initial data support for generating a final evaluation by combining more complex logic (such as expert experience). The basic scoring results include, but are not limited to, the score value of at least one basic scoring indicator and the corresponding score explanation information for each basic scoring indicator.

[0054] Specifically, the target search results are input into the rule engine, and matched according to the predefined scoring rules in the rule engine. The rule engine will automatically perform quantitative calculations and logical judgments on various information in the target search results. For example, if the registered capital of an enterprise reaches a certain threshold, the corresponding score will be added; if the data of a certain dimension does not meet the basic requirements, the corresponding score will be deducted. Finally, the basic score result is obtained by summarizing the various calculation results and the scoring rules and related information corresponding to each result.

[0055] In this embodiment, the basic scoring results are determined by a rule engine, which is clear and objective, avoids the subjective bias of manual scoring, and can achieve standardized scoring based on specific data in the target retrieval results. At the same time, the automated calculation process greatly improves scoring efficiency and can quickly output consistent basic scoring results, providing a fair and reliable quantitative basis for subsequent score-based decision-making.

[0056] Optionally, determining the basic scoring result based on the target retrieval result using a rule engine includes: calling the rule engine and loading basic scoring rules, wherein the basic scoring rules include the correspondence between dimensions and scoring rules; inputting the target retrieval result into the rule engine, parsing the target retrieval result through the rule engine, and extracting data details for at least one dimension; matching at least one dimension with the basic scoring rules, and determining the scoring result for each dimension based on the successfully matched scoring rules and the data details for each dimension; and performing fusion processing based on the scoring results for each dimension to determine the basic scoring result, wherein the basic scoring result includes the score value of at least one basic scoring indicator and the scoring explanation information corresponding to the score value of each basic scoring indicator.

[0057] The basic score result is a quantitative evaluation result obtained by scoring and integrating data from various dimensions based on the target search results through a rule engine. It includes the specific score value of at least one basic score indicator, which is a comprehensive quantitative reflection of the company's performance in various dimensions under a specific evaluation scenario, providing standardized basic values ​​for subsequent analysis or decision-making. The score explanation information is the explanatory content corresponding to the score value of each basic score indicator. It explains in detail the origin of the score value. For example, a score for a certain indicator is given as a bonus because the company has been established for 5 years or more, or a score is deducted because the data in a certain dimension does not meet the standard. This allows users to clearly understand the basis of the score and enhances the transparency and credibility of the basic score result.

[0058] Specifically, the process begins by invoking the rule engine and loading basic scoring rules. These basic scoring rules define the scoring dimensions and their corresponding scoring rules, meaning they contain a clear correspondence between each dimension and its corresponding scoring rule. Next, the target retrieval results are input into the rule engine, which then parses the results and extracts data details for at least one dimension. Subsequently, the rule engine matches the extracted dimensions with the basic scoring rules, calculating the score for each dimension based on the successfully matched rules and the specific data details for each dimension. Finally, the scoring results for each dimension are fused to generate a basic scoring result containing at least one basic scoring indicator and corresponding explanation information for each score.

[0059] In this embodiment, the pre-loading of basic scoring rules ensures the standardization and consistency of the scoring logic. The rule engine's accurate parsing and dimension matching of the target retrieval results guarantee the relevance and accuracy of the scoring. At the same time, the clear presentation of scores and explanatory information for each dimension during the scoring process enhances the traceability and interpretability of the basic scoring results. This provides a transparent and reliable quantitative basis for generating a more comprehensive evaluation by combining expert scoring rules, and also improves the efficiency and standardization of the entire scoring process.

[0060] S140. Obtain expert scoring rules. Based on the expert scoring rules, target retrieval results, and basic scoring results, generate a target enterprise profile using a pre-built enterprise profile generation model. The target enterprise profile is generated according to a preset enterprise profile template. The target enterprise profile includes the total score, the score value of at least one scoring indicator, and the scoring explanation information corresponding to each score value.

[0061] Among them, expert scoring rules are rules developed by domain experts based on their professional experience and business needs to supplement or optimize scoring logic. These rules include subjective judgment standards or refined scoring criteria for special scenarios and complex dimensions, providing professional-level supplementation to the scoring. Expert scoring rules developed by domain experts based on their experience and professional knowledge can be collected and organized first, stored in a designated storage space and updated in real time. When applying the application, the latest expert scoring rules can be retrieved from the designated storage space. The enterprise profile generation model can be understood as an intelligent model built on large language model technology. Relying on semantic understanding and generation capabilities formed through training on massive amounts of text data, it can receive input information such as expert scoring rules, target retrieval results, and basic scoring results. Through deep analysis of the semantic relationships of this information, combined with a preset enterprise profile template, it uses natural language generation capabilities to structurally integrate and optimize the expression of information, ultimately generating a target enterprise profile that conforms to the template specifications. This model can accurately integrate objective data and professional rules, ensuring a standardized and unified output format while endowing the profile with rich details and clear interpretability. It efficiently realizes the transformation from scattered information to a complete enterprise profile, providing users with decision support that is both professional and practical. A target company profile can be understood as a structured presentation reflecting the overall situation of the target company, including the total score, the score values ​​of each scoring indicator, and the corresponding explanatory information. It can be generated by a pre-built company profile generation model based on a preset company profile template. The preset company profile template is a pre-defined profile framework that specifies the content structure, presentation format, and included elements of the profile, such as the categories of scoring indicators and the format of explanatory information, ensuring that the generated company profiles are uniform and standardized.

[0062] Specifically, the latest expert scoring rules are retrieved from the designated storage space. These rules, along with various detailed information about the company in the target search results, the score values ​​and explanations for each dimension in the basic scoring results, are then input into the pre-built company profile generation model. The pre-built company profile generation model integrates the professional judgment of the expert scoring rules, the raw data support of the target search results, and the quantitative basis of the basic scoring results. It performs structured processing according to the preset company profile template, and finally generates a target company profile that includes the total score, the score value of at least one scoring indicator, and the score explanation information corresponding to each score value.

[0063] In this embodiment, the integration of expert scoring rules ensures that the profile possesses both data objectivity and professional depth, avoiding the one-sidedness of purely data-driven approaches. Relying on target retrieval results and basic scoring results ensures the authenticity of the profile content and the reliability of the quantitative basis. The structured profile generated according to the preset template includes the total score, specific scores, and explanatory information, which not only intuitively presents the overall situation of the enterprise but also allows users to clearly understand the scoring logic, thereby improving the readability and practical value of the profile and providing comprehensive and professional reference for enterprise evaluation, cooperation decision-making, and other scenarios.

[0064] Based on the above embodiments, the method further includes: verifying the basic scoring results based on a first preset verification rule; if the basic scoring results do not meet the first preset verification conditions, invoking a rule engine optimization processing method to optimize the rule engine, wherein the first preset verification rule includes integrity verification rules and / or consistency verification rules; and / or, verifying the target enterprise profile based on a second preset verification rule; if the target enterprise profile does not meet the second preset verification conditions, invoking an enterprise profile generation model optimization processing method to optimize the pre-built enterprise profile generation model, wherein the second preset verification rule includes enterprise profile information integrity verification rules.

[0065] The first preset verification rule can be understood as a standard used to verify the basic scoring results, including completeness verification rules and / or consistency verification rules. The completeness verification rules check whether the basic scoring results cover all necessary basic scoring indicators, while the consistency verification rules verify whether the scoring logic of each dimension is consistent and without contradiction, thereby ensuring the compliance of the basic scoring results in terms of content and logic. The second preset verification rule can be understood as a verification standard for the target enterprise profile, with its core being the enterprise profile information completeness verification rule. This includes checking whether the profile omits one or more of the necessary content such as the total score, the score values ​​of key scoring indicators, and corresponding explanatory information, to ensure the completeness and standardization of the target enterprise profile. The enterprise profile generation model optimization processing method is the adjustment method used when the target enterprise profile does not meet the second preset verification conditions. This includes optimizing the model's integration logic of expert scoring rules, target retrieval results, and basic scoring results, or improving the preset enterprise profile template elements called by the model. The aim is to improve the model's ability to generate target enterprise profiles that meet the verification requirements through targeted adjustments.

[0066] Specifically, based on a first preset verification rule that includes integrity and / or consistency verification rules, a compliance check is performed on the basic scoring results. If the verification result does not meet the verification conditions, the rule engine optimization processing method is invoked to optimize the rule engine. For example, the rule engine optimization processing method includes adjusting the weight of the scoring rules and / or supplementing the scoring criteria for missing dimensions. The specific optimization method is set according to the actual situation and is not limited here. Alternatively, the target enterprise profile is verified based on a second preset verification rule. If the verification result does not meet the verification conditions, the enterprise profile generation model optimization processing method is invoked to optimize the model. The enterprise profile generation model optimization processing method includes, but is not limited to, adjusting the information fusion logic of the model and supplementing the template mandatory field verification mechanism. The above two verification mechanisms can also be triggered simultaneously to ensure the accuracy of the target enterprise profile.

[0067] In this embodiment, the dual verification mechanism can promptly detect potential problems in the basic scoring results and target enterprise profiles. Through targeted optimization, it ensures the output quality of the rule engine and model, which not only guarantees the rigor of the basic scoring and the integrity of the enterprise profiles, but also enables the system to self-iterate and upgrade, thereby improving the reliability and adaptability of the overall process.

[0068] Based on the above embodiments, the method further includes: obtaining timestamp information of multiple data sources corresponding to the pre-built enterprise knowledge base; if the timestamp information of any data source is not the latest publication time, obtaining the latest published content of the data source; and updating the pre-built enterprise knowledge base based on the latest published content.

[0069] Specifically, firstly, the system connects to multiple data sources associated with the enterprise knowledge base via a pre-defined data acquisition interface, automatically retrieving timestamp information from each data source. These timestamps record the content's publication or update time. Simultaneously, a scheduled task tool (such as CRON) triggers the timestamp retrieval process at a pre-defined frequency (e.g., daily at midnight) to ensure timely and automated information collection. Next, the timestamp information from each data source is compared with the latest publication time published through official channels to determine if any data source has outdated timestamps. If any data source's timestamp is found to be outdated, its latest published content is immediately retrieved via the data source's open interface or authorized crawler technology. Finally, based on the retrieved latest published content, an incremental update mechanism precisely replaces or supplements the corresponding data entries in the enterprise knowledge base. This involves updating the pre-built enterprise knowledge base with the newly acquired content according to its format and rules, completing the knowledge update iteration and avoiding the resource consumption of a full update. This ensures that the information related to the data source in the knowledge base remains consistent with the latest published content, thereby efficiently maintaining the timeliness and accuracy of the enterprise knowledge base.

[0070] In this embodiment, the knowledge in the enterprise knowledge base is ensured to remain timely and accurate, allowing employees and relevant users to access the latest industry trends, policies and regulations, technological developments, and other information, providing strong support for enterprise decision-making and business operations. Simultaneously, the automated update process reduces manual intervention, improves the efficiency of knowledge management, reduces the possibility of human error, and helps break down knowledge silos, integrating the latest knowledge scattered across various locations into a unified knowledge base, thereby enhancing the value of the enterprise's knowledge assets.

[0071] Optionally, to facilitate the querying and generation of target enterprise profiles, the visualization platform can provide functions such as an enterprise search interface, an enterprise profile display page, visualization of scoring results (radar chart display of sub-scores), and explanatory text download. The processing of input information and the generation of target enterprise profiles can be completed through the backend. After the backend generates the target enterprise profile, it is pushed to the frontend for display.

[0072] Optionally, after obtaining the target company profile, the company profile and intelligent scoring algorithm can be called to score the company profile. If the score meets the preset threshold, the target company profile is output. If the score does not meet the preset threshold, the target company profile generation method is called again to obtain a new target company profile, until the score of the obtained target company profile meets the preset threshold, so as to improve the accuracy and effectiveness of the target company profile.

[0073] The technical solution of this embodiment involves acquiring input information, performing semantic parsing on the input information to obtain semantic information, wherein the semantic information includes the company name and intent type; searching in a pre-built enterprise knowledge base based on the company name and intent type to obtain target search results; determining basic scoring results based on the target search results through a rule engine; acquiring expert scoring rules; and generating a target company profile based on the expert scoring rules, the target search results, and the basic scoring results through a pre-built enterprise profile generation model, wherein the target company profile is generated according to a preset company profile template and includes a total score, a score value for at least one scoring indicator, and scoring explanation information corresponding to each score value. This solution accurately extracts company names and intent types through semantic parsing of input information, ensuring the relevance and accuracy of subsequent searches. Relying on a pre-built company knowledge base provides comprehensive and reliable basic data support for scoring and profile generation. The basic scoring results determined by the rule engine guarantee the objectivity and consistency of the scoring, while incorporating expert scoring rules supplements professional experience judgment, making the scoring both data rigor and domain depth. The company profile generation model generates target company profiles containing total scores, specific scores, and corresponding explanations according to preset templates. This achieves both structured and standardized profiles, and improves readability through clear scoring logic, solving the problems of not being able to obtain effective company profiles and poor interpretability. The entire process forms a closed loop from information input to result output, balancing automation efficiency and professional reliability, providing comprehensive, accurate, and easy-to-understand references for relevant corporate decisions.

[0074] Example 2

[0075] Figure 2 This is a flowchart of a method for generating enterprise profiles according to Embodiment 2 of the present invention. The method in this embodiment is a further optimization of the method in the above embodiments. Optionally, a pre-built enterprise profile generation prompt template is obtained; enterprise profile generation prompt information is determined based on expert scoring rules, basic scoring results, and the enterprise profile generation prompt template; the enterprise profile generation prompt information and target retrieval results are input into a pre-built enterprise profile generation model, and the pre-built enterprise profile generation model outputs the target enterprise profile. Figure 2 As shown, the method includes:

[0076] S210. Obtain input information, perform semantic parsing on the input information to obtain semantic information, which includes the company name and intent type.

[0077] S220. Based on the company name and intent type, perform a search in the pre-built company knowledge base to obtain the target search results.

[0078] S230. Determine the basic scoring results based on the target retrieval results using a rule engine.

[0079] S240. Obtain expert scoring rules, obtain pre-built enterprise profile generation prompt templates, and determine enterprise profile generation prompt information based on expert scoring rules, basic scoring results, and enterprise profile generation prompt templates.

[0080] The enterprise profile generation prompt template can be understood as a framework used to standardize the structure and core elements of enterprise profile generation prompts. It unifies and standardizes the generation direction and content scope of enterprise profiles, providing a basic framework for subsequent information filling. The enterprise profile generation prompt template should at least include the generation objective, scoring rule references, and format requirements. For example, based on [expert scoring rules] and [basic scoring results], it generates fixed expressions such as enterprise profiles covering [total score], [scoring of each indicator], and [corresponding explanations], as well as placeholders for filling in specific content. This provides a unified standard for prompt information generation and can be adjusted according to actual needs. Specifically, the enterprise profile generation prompt information refers to the specific prompt content formed by integrating key data such as expert scoring rules, basic scoring results, and target retrieval results from the enterprise profile generation prompt template. It clearly guides the enterprise profile generation model to generate a standardized target enterprise profile according to preset requirements.

[0081] Specifically, the process first retrieves a pre-built enterprise profile generation prompt template containing the core elements required for profile generation. This template specifies the structure and key content direction of the prompt information. Then, key data such as expert scoring rules (e.g., scoring standards for special scenarios) and basic scoring results (including the scoring values ​​and explanatory information of each indicator) are filled in and integrated according to the template's format requirements. This clarifies the expert judgment criteria, basic quantitative results, and related explanatory logic that need to be referenced when generating the enterprise profile, ultimately forming a complete enterprise profile generation prompt information.

[0082] In this embodiment, using a pre-built enterprise profile generation prompt template can ensure the standardization and completeness of the prompt information and avoid the omission of key elements. Combining expert scoring rules and basic scoring results, the prompt information includes both professional experience guidance and quantitative data support, providing clear and accurate guidance for the enterprise profile generation model. This helps the model generate target enterprise profiles that are more in line with needs and are both professional and objective. At the same time, it improves the efficiency and consistency of the profile generation process and reduces the redundant costs of manual writing.

[0083] S250. Input the enterprise profile generation prompt information and target search results into the pre-built enterprise profile generation model, and the pre-built enterprise profile generation model outputs the target enterprise profile.

[0084] Specifically, the structured enterprise profile generation prompts and target search results are input into a pre-built enterprise profile generation model. The pre-built enterprise profile generation model first parses the core instructions in the prompts, clarifying the elements and format specifications that the profile must include, such as the total score, scoring indicators, and explanatory information. Then, it combines the specific data in the target search results and uses semantic association analysis to match the data with the rules and indicators in the prompts (for example, matching "annual revenue growth of 15%" to the "operational capability score" dimension). Subsequently, based on the text generation capabilities of the large language model, the model integrates the data and rule logic according to the prompt requirements to generate a target enterprise profile that conforms to the preset template, including the total score, the specific scores of each scoring indicator, and the corresponding detailed explanations. The detailed explanations include the data source and scoring basis.

[0085] In this embodiment, the enterprise profile generation prompts provide clear generation guidelines for the enterprise profile generation model, ensuring that the enterprise profile format is uniform and the elements are complete. The target retrieval results provide a real and accurate enterprise data foundation, ensuring the objectivity of the profile content. The combination of these two inputs into the enterprise profile generation model ensures that the output target enterprise profile not only strictly follows the generation requirements but also deeply integrates actual enterprise information, taking into account standardization, authenticity, and interpretability. At the same time, the model's natural language processing capabilities make the profile description fluent and easy to understand, significantly improving the quality and efficiency of enterprise profile generation.

[0086] The technical solution of this embodiment obtains input information, performs semantic parsing on the input information to obtain semantic information, including the company name and intent type; searches a pre-built enterprise knowledge base based on the company name and intent type to obtain target search results; determines basic scoring results based on the target search results using a rule engine; obtains expert scoring rules; obtains a pre-built enterprise profile generation prompt template; and determines enterprise profile generation prompt information based on the expert scoring rules, basic scoring results, and enterprise profile generation prompt template; inputs the enterprise profile generation prompt information and the target search results into a pre-built enterprise profile generation model, which then outputs the target enterprise profile. This solution forms a complete closed loop from information input to profile output. Semantic parsing ensures the accuracy of core element extraction, the enterprise knowledge base provides a reliable data foundation for retrieval, the rule engine ensures the objectivity of the basic scoring, expert scoring rules incorporate professional judgment, the prompt template standardizes the generation direction, and the model efficiently integrates information to generate a structured profile. The overall process balances automation efficiency and professional depth, ensuring that the output target enterprise profile is accurate, complete, and practical, providing strong support for enterprise decision-making and other scenarios.

[0087] Example 3

[0088] Figure 3This is a schematic diagram of the structure of an enterprise profile generation device provided in Embodiment 3 of the present invention.

[0089] like Figure 3 As shown, the device includes:

[0090] The semantic information determination module 310 is used to acquire input information, perform semantic parsing on the input information, and obtain semantic information, wherein the semantic information includes the enterprise name and intent type;

[0091] The target retrieval result determination module 320 is used to perform a retrieval in a pre-built enterprise knowledge base based on the enterprise name and intent type to obtain the target retrieval results;

[0092] The basic scoring result determination module 330 is used to determine the basic scoring result based on the target retrieval result through a rule engine.

[0093] The target enterprise profile generation module 340 is used to obtain expert scoring rules and generate a target enterprise profile based on the expert scoring rules, target retrieval results and basic scoring results through a pre-built enterprise profile generation model. The target enterprise profile is generated according to a preset enterprise profile template and includes a total score, a score value of at least one scoring indicator and a score explanation information corresponding to each score value.

[0094] The technical solution of this embodiment involves a semantic information determination module acquiring input information, performing semantic parsing on the input information to obtain semantic information, which includes the company name and intent type; a target retrieval result determination module retrieving from a pre-built enterprise knowledge base based on the company name and intent type to obtain target retrieval results; a basic scoring result determination module determining basic scoring results based on the target retrieval results using a rule engine; and a target enterprise profile generation module acquiring expert scoring rules, generating a target enterprise profile based on the expert scoring rules, the target retrieval results, and the basic scoring results using a pre-built enterprise profile generation model. The target enterprise profile is generated according to a preset enterprise profile template and includes a total score, a score value for at least one scoring indicator, and scoring explanation information corresponding to each score value. This solution accurately extracts company names and intent types through semantic parsing of input information, ensuring the relevance and accuracy of subsequent searches. Relying on a pre-built company knowledge base provides comprehensive and reliable basic data support for scoring and profile generation. The basic scoring results determined by the rule engine guarantee the objectivity and consistency of the scoring, while incorporating expert scoring rules supplements professional experience judgment, making the scoring both data rigor and domain depth. The company profile generation model generates target company profiles containing total scores, specific scores, and corresponding explanations according to preset templates. This achieves both structured and standardized profiles, and improves readability through clear scoring logic, solving the problems of not being able to obtain effective company profiles and poor interpretability. The entire process forms a closed loop from information input to result output, balancing automation efficiency and professional reliability, providing comprehensive, accurate, and easy-to-understand references for relevant corporate decisions.

[0095] Based on the above embodiments, optionally, the semantic information determination module 310 is specifically used to preprocess the input information to obtain preprocessed data, wherein the preprocessing includes text cleaning, word segmentation and stop word filtering; extract the enterprise name and key fields from the preprocessed data; and determine the intent type corresponding to the input information based on the key fields through a pre-built intent recognition model, wherein the intent type is a type set based on demand information, and the demand information includes user attention dimension, information depth and application scenario.

[0096] Optionally, the target retrieval result determination module 320 is specifically used to search the business registration information in the enterprise knowledge base based on the enterprise name, determine the enterprise identifier and basic enterprise information corresponding to the enterprise name; call the intent and dimension mapping interface based on the intent type to obtain at least one dimension corresponding to the intent type; search the enterprise knowledge base based on the enterprise identifier and at least one dimension to determine the data details corresponding to each dimension; and perform fusion processing based on the enterprise identifier, basic enterprise information, at least one dimension and the data details corresponding to each dimension to obtain the target retrieval result.

[0097] Optionally, the basic scoring result determination module 330 is specifically used to call the rule engine, load basic scoring rules, wherein the basic scoring rules include the correspondence between dimensions and scoring rules; input the target retrieval results into the rule engine, and parse the target retrieval results through the rule engine to extract data details of at least one dimension; match at least one dimension with the basic scoring rules, and determine the scoring results of each dimension based on the successfully matched scoring rules and the data details of each dimension; perform fusion processing based on the scoring results of each dimension to determine the basic scoring results, wherein the basic scoring results include the scoring value of at least one basic scoring indicator and the scoring explanation information corresponding to the scoring value of each basic scoring indicator.

[0098] Optionally, the target enterprise profile generation module 340 is specifically used to obtain a pre-built enterprise profile generation prompt template, determine enterprise profile generation prompt information based on expert scoring rules, basic scoring results and the enterprise profile generation prompt template, input the enterprise profile generation prompt information and target search results into the pre-built enterprise profile generation model, and output the target enterprise profile by the pre-built enterprise profile generation model.

[0099] Optionally, the device is further configured to verify the basic scoring results based on a first preset verification rule; if the basic scoring results do not meet the first preset verification condition, the device calls a rule engine optimization processing method to optimize the rule engine, wherein the first preset verification rule includes integrity verification rules and / or consistency verification rules; and / or, to verify the target enterprise profile based on a second preset verification rule; if the target enterprise profile does not meet the second preset verification condition, the device calls an enterprise profile generation model optimization processing method to optimize the pre-built enterprise profile generation model, wherein the second preset verification rule includes enterprise profile information integrity verification rules.

[0100] Optionally, the device is also used to obtain timestamp information from multiple data sources corresponding to the pre-built enterprise knowledge base. If the timestamp information of any data source is not the latest publication time, the device obtains the latest published content from the data source and updates the pre-built enterprise knowledge base based on the latest published content.

[0101] The enterprise profile generation device provided in this embodiment of the invention can execute the enterprise profile generation method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0102] Example 4

[0103] Figure 4This is a schematic diagram of the structure of an electronic device provided in Embodiment 4 of the present invention. The electronic device 10 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0104] like Figure 4 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 may also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0105] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0106] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as enterprise profiling generation methods.

[0107] In some embodiments, the enterprise profile generation method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the enterprise profile generation method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the enterprise profile generation method by any other suitable means (e.g., by means of firmware).

[0108] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0109] Computer programs used to implement the enterprise profile generation method of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The computer programs can be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0110] Example 5

[0111] Embodiment 5 of the present invention also provides a computer-readable storage medium storing computer instructions for causing a processor to execute a method for generating an enterprise profile, the method comprising:

[0112] Obtain input information, perform semantic parsing on the input information to obtain semantic information, which includes the company name and intent type;

[0113] The search results are obtained by performing a search on a pre-built enterprise knowledge base based on the enterprise name and intent type.

[0114] The basic scoring results are determined using a rule engine based on the target search results;

[0115] The system obtains expert scoring rules, and generates a target enterprise profile based on the expert scoring rules, target search results, and basic scoring results using a pre-built enterprise profile generation model. The target enterprise profile is generated according to a preset enterprise profile template and includes a total score, a score value for at least one scoring indicator, and a score explanation for each score value.

[0116] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0117] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0118] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0119] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through a communication network. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0120] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0121] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for generating enterprise profiles, characterized in that, include: Obtain input information, perform semantic parsing on the input information to obtain semantic information, wherein the semantic information includes the company name and intent type; Based on the enterprise name and intent type, a search is performed in a pre-built enterprise knowledge base to obtain the target search results; Based on the target retrieval results, a basic scoring result is determined using a rule engine; Obtain expert scoring rules, and generate a target enterprise profile based on the expert scoring rules, the target retrieval results, and the basic scoring results using a pre-built enterprise profile generation model. The target enterprise profile is generated according to a preset enterprise profile template and includes a total score, a score value for at least one scoring indicator, and scoring explanation information corresponding to each score value.

2. The method according to claim 1, characterized in that, The step of performing semantic parsing on the input information to obtain semantic information includes: The input information is preprocessed to obtain preprocessed data, wherein the preprocessing includes text cleaning, word segmentation, and stop word filtering; Extract the company name and key fields from the preprocessed data; Based on the key fields, the intent type corresponding to the input information is determined by a pre-built intent recognition model. The intent type is a type set based on demand information, which includes user attention dimensions, information depth, and application scenarios.

3. The method according to claim 1, characterized in that, The process of retrieving the target search results from a pre-built enterprise knowledge base based on the enterprise name and intent type includes: Based on the enterprise name, a search is performed in the business registration information in the enterprise knowledge base to determine the enterprise identifier and basic enterprise information corresponding to the enterprise name; Based on the intent type, the intent-dimensional mapping interface is invoked to obtain at least one dimension corresponding to the intent type; Based on the enterprise identifier and the at least one dimension, a search is performed in the enterprise knowledge base to determine the data details corresponding to each dimension. The target retrieval result is obtained by fusing the enterprise identifier, the enterprise basic information, the at least one dimension, and the data details corresponding to each dimension.

4. The method according to claim 1, characterized in that, The determination of the basic scoring result based on the target retrieval result using a rule engine includes: The rules engine is invoked to load the basic scoring rules, wherein the basic scoring rules include the correspondence between dimensions and scoring rules; The target retrieval results are input into the rule engine, which then parses the results to extract data details from at least one dimension. The at least one dimension is matched with the basic scoring rule, and the scoring result of each dimension is determined based on the successfully matched scoring rule and the data details of each dimension; The basic scoring result is determined by fusing the scoring results of each dimension. The basic scoring result includes the scoring value of at least one basic scoring indicator and the scoring explanation information corresponding to the scoring value of each basic scoring indicator.

5. The method according to claim 1, characterized in that, The process of generating a target enterprise profile based on the expert scoring rules, the target retrieval results, and the basic scoring results using a pre-built enterprise profile generation model includes: Obtain a pre-built enterprise profile generation prompt template, and determine enterprise profile generation prompt information based on the expert scoring rules, the basic scoring results, and the enterprise profile generation prompt template; The enterprise profile generation prompt information and the target search results are input into the pre-built enterprise profile generation model, and the pre-built enterprise profile generation model outputs the target enterprise profile.

6. The method according to claim 1, characterized in that, The method also includes: The basic scoring results are verified based on a first preset verification rule. If the basic scoring results do not meet the first preset verification condition, a rule engine optimization processing method is invoked to optimize the rule engine. The first preset verification rule includes integrity verification rules and / or consistency verification rules; and / or... The target enterprise profile is verified based on the second preset verification rule. If the target enterprise profile does not meet the second preset verification condition, the enterprise profile generation model optimization processing method is invoked to optimize the pre-built enterprise profile generation model. The second preset verification rule includes the enterprise profile information integrity verification rule.

7. The method according to claim 1, characterized in that, The method also includes: Obtain timestamp information from multiple data sources corresponding to the pre-built enterprise knowledge base. If the timestamp information of any of the data sources is not the latest publication time, obtain the latest published content of the data source and update the pre-built enterprise knowledge base based on the latest published content.

8. A corporate profile generation device, characterized in that, include: The semantic information determination module is used to acquire input information, perform semantic parsing on the input information, and obtain semantic information, wherein the semantic information includes the enterprise name and intent type; The target retrieval result determination module is used to perform a retrieval in a pre-built enterprise knowledge base based on the enterprise name and intent type to obtain the target retrieval result; The basic scoring result determination module is used to determine the basic scoring result based on the target retrieval result through a rule engine; The target enterprise profile generation module is used to obtain expert scoring rules, and generate a target enterprise profile based on the expert scoring rules, the target retrieval results and the basic scoring results through a pre-built enterprise profile generation model. The target enterprise profile is generated according to a preset enterprise profile template, and the target enterprise profile includes a total score, a score value of at least one scoring indicator and a score explanation information corresponding to each score value.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the enterprise profile generation method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the enterprise profile generation method according to any one of claims 1-7.

Images