Commercial information processing method

Abstract

The invention provides a business information processing method, which comprises the following steps: acquiring business card information input by a user, and generating first business card data of the user through an artificial intelligence model; obtaining business project demand data input by the user; generating business project scheme data according to the business project demand data; matching at least one piece of second business card data through a business activity artificial intelligence model according to the business project scheme data, and forming a second business card data set corresponding to the business project demand of the user; and establishing business information processing between the first business card data and the second business card data according to the second business card data set. According to the invention, discovery, communication and commercial activity cooperation of commercial contacts are realized more efficiently and more accurately.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a method for processing business information. Background Art

[0002] Currently, traditional business cards are mainly in paper form, with fixed and limited information displayed, high production and distribution costs, and a lack of personalized customization and intelligent functions. In terms of business social networking, common social platforms focus on maintaining personal social relationships and lack the accuracy to expand business connections. Their methods of finding potential business partners mainly rely on manual searches by users and offline activities, which is inefficient. In terms of business cooperation matching, partners are usually screened based on manual experience, which makes it difficult to comprehensively and accurately match the business capabilities and needs of both parties, and it is also impossible to automatically generate feasible business cooperation plans. For example, in some traditional business social activities, after participants exchange paper business cards, it is difficult to quickly find partners that truly match their business needs. It often takes a lot of time and energy to communicate and screen.

[0003] Among the existing technologies, such as patents, one patent, titled "A Business Promotion Model Based on Electronic Business Cards" (application number CN201710784403.3), provides an implementation of a business promotion model based on electronic business cards. The patent proposes an electronic business card-based business promotion model that utilizes an internet website and mobile terminals to create an e-commerce online trading platform app. Enterprise users A and B open corporate accounts through the app to become registered companies, manage their business card information, and open an online marketplace to publish their product information. Employees of user A and user B register as business card users under their respective company's business cards. Each employee has their own business card space. Users A and B can negotiate to establish a small broadcast alliance to share their company's product information. This electronic business card-based business promotion model offers the following benefits: leveraging the personal connections and social resources of enterprise employees to achieve the most cost-effective and efficient promotion model for enterprises. Registered enterprises can establish small broadcast alliances to enable employees from different companies to exchange and share their social resources.

[0004] With the rise of AI (Artificial Intelligence) technology, there is a trend of combining AI big model technology with business information provision technology. Commercial companies are also training big models based on their own company data to provide users with a more tailored business information experience.

[0005] In summary, in the existing technologies, there is no use of artificial intelligence technology in the generation and use of business information, and no solutions are provided for the efficient use of business information. Summary of the Invention

[0006] The embodiment of the present invention provides a method for processing business information, which is used to implement a business information model based on artificial intelligence technology to generate business user business cards and build accurate and efficient business network circles, and can promote and implement business activities accurately and efficiently.

[0007] In a first aspect, an embodiment of the present invention provides a method for processing business information, including: obtaining business card information input by a user, and generating first business card data of the user through an artificial intelligence model; obtaining business project demand data input by the user; generating business project plan data based on the business project demand data; matching at least one second business card data through a business activity artificial intelligence model according to the business project plan data to form a second business card data set corresponding to the user's business project demand; and establishing business information processing between the first business card data and the second business card data based on the second business card data set.

[0008] Optionally, the above method further includes: matching at least one third business card data based on the user's capability information and / or user's business resource information and / or business product / service information provided by the user and / or user demand information in the first business card data to form a third business card data set corresponding to the first business card data; and establishing business information processing between the first business card data and the third business card data based on the third business card data set.

[0009] Optionally, matching at least one third business card data includes: obtaining business card data, user business friend information, user behavior data, and user chat content data; analyzing and generating a social relationship graph based on a business social network analysis model; generating a user interest vector based on a user behavior model; and generating matching at least one third business card data based on the social relationship graph and the user interest vector in a hybrid recommendation model.

[0010] Optionally, the obtaining of business card information input by the user and generating the first business card data of the user through an artificial intelligence model includes: obtaining user identity information and / or business card style information and / or user capability information and / or user business resource information and / or business product / service information provided by the user and / or user demand information input by the user; calling a text-based artificial intelligence model to generate text for the user's first business card; calling a picture-based artificial intelligence model to generate a picture of the user's first business card; and constructing the first business card data of the user based on the text in the generated first business card and the picture in the first business card.

[0011] Optionally, generating business project solution data based on the business project demand data includes: obtaining descriptive information of the business project demand; calling a business project implementation model to generate a business solution sub-step for realizing the business project demand; in the business solution sub-step, generating second business card data and / or business cooperation solution data corresponding to the business solution sub-step based on the user's capability information and / or the user's business resource information and / or the business product / service information provided by the user.

[0012] Optionally, the above method further includes: obtaining corporate business card information input by the corporate user to generate first corporate business card data of the corporate user; wherein the first corporate business card data is associated with at least one first business card data; obtaining corporate knowledge base training data of the corporate user to generate an enterprise product and service knowledge base model; and responding to product and service demand information of the second business card data based on the enterprise product and service knowledge base model.

[0013] Optionally, responding to the product and service demand information of the second business card data based on the enterprise product and service knowledge base model includes: generating a business plan sub-step for realizing the product and service demand of the second business card data based on the product and service demand of the second business card data as input; using business cases and / or industry rules in the knowledge base model and generating business cooperation plan data corresponding to the business plan sub-step based on the business plan sub-step as input.

[0014] Optionally, the above method further includes: in the business information processing between the first business card data and the second business card data, loading the business contract data generated by the contract macro model based on the business project demand data and the business project plan data; editing and signing the business contract data based on the first business card data and the second business card data.

[0015] Optionally, matching at least one second business card data through a business activity artificial intelligence model according to the business project plan data to form a second business card data set corresponding to the user's business project needs includes: performing semantic parsing on the business project plan data through a business activity artificial intelligence model to generate a structured semantic architecture; extracting domain entities from the structured semantic architecture; constructing query conditions based on the extracted domain entities, and retrieving a second business card data candidate set that meets the conditions from the business activity knowledge graph.

[0016] Optionally, the above method further includes: collecting data from business card data and / or corporate business information and / or product and service description information and / or user social content information and / or business contract transaction information; performing mixed extraction of business activity knowledge; and integrating the business activity knowledge of upstream and downstream business activity entities to form a global business activity knowledge graph.

[0017] The present invention uses an artificial intelligence model to generate the user's business card data from input business card information. It then translates the user's business project requirements into business project proposals. Using a business activity artificial intelligence model, it matches at least one business card data point to meet the user's business project requirements. This then forms a business network with the matching business cards, facilitating efficient and accurate information exchange between business activity users within the network, ultimately achieving business goals. This approach, fundamentally different from traditional business card-based business activity communication, allows for more efficient and accurate discovery, exchange, and collaboration on business activities. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following is a brief introduction to the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1 A flowchart of a method for processing commercial information provided in the first embodiment of the present invention; Figure 2 A schematic diagram of the framework principle of a commercial information processing system provided in the second embodiment of the present invention; Figure 3 A schematic diagram of a process for generating and processing a business project plan in a business information processing system provided in the third embodiment of the present invention; Figure 4 A schematic diagram of a process for business information processing by an enterprise user according to the fourth embodiment of the present invention; Figure 5Screenshot of the APP interface for processing business demand input by individual users provided in Example 5 of the present invention; Figure 6 This is a screenshot of the APP interface for analyzing business needs, generating solutions, and matching business cards with partners, as provided in Example 5 of the present invention. DETAILED DESCRIPTION

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0021] The terms used in the embodiments of the present invention are for the purpose of describing specific embodiments only and are not intended to limit the present invention. The singular forms "a," "an," "the," and "the" used in the embodiments of the present invention and the appended claims are also intended to include plural forms, and unless the context clearly indicates otherwise, "a plurality" generally includes at least two.

[0022] As used herein, the words "if" and "if" may be interpreted as "at the time of" or "when" or "in response to determining" or "in response to detecting," depending on the context. Similarly, the phrases "if it is determined" or "if (stated condition or event) is detected" may be interpreted as "when it is determined" or "in response to the determination" or "when detecting (stated condition or event)" or "in response to detecting (stated condition or event)," depending on the context.

[0023] In addition, the step sequence in the following method embodiments is only an example and not a strict limitation.

[0024] In an embodiment of the present invention, an AI model generates a user's business card data, converts the user's business project requirements into business proposals, and matches at least one business card data item with the user's business project requirements using an AI business activity model. This information then forms a business network with the matching business card, facilitating efficient and accurate information exchange between business activity users within the network to achieve business goals. This method, in contrast to traditional business card-based business activity communication, enables more efficient and accurate discovery, exchange, and collaboration on business activities.

[0025] Example 1: Figure 1This is a flowchart of a method for processing commercial information provided in Example 1 of the present invention. The method for processing commercial information in Example 1 includes the following steps: Step 100: Obtain the business card information input by the user and generate the first business card data of the user through the artificial intelligence model In this embodiment, a user enters business card information through the client interface, specifically including user identity information (such as name "Zhang San", position "Marketing Manager", and contact information "138xxxx1234"), business card style information (user selects "Simple Business Style"), user capability information (marked as "5 years of marketing experience, proficient in digital marketing"), user business resource information (such as "30+ media partners"), user-provided business product / service information ("Providing full-case corporate brand planning services"), and user needs information ("Seeking brand promotion projects with an annual budget of over 1 million yuan"). The system then invokes a text-based artificial intelligence model (such as ChatGPT-4) to generate business card text based on the structured information input. For example: Zhang San, Marketing Manager at XX Technology Co., Ltd., 5 years of marketing experience | Digital Marketing Expert. Services: Full-scale brand planning, social media operations, and partnership consulting. Contact: 138xxxx1234. Simultaneously, an image-based AI model (such as DALL-E) is invoked to generate a business card background image (a light gray gradient background with the company logo) based on the "minimalist business style" directive. Finally, the generated text content is formatted and integrated with the image (using a left-right column layout, with the image on the left and the text on the right), creating the first business card data, which contains structured data (name, position, contact information, etc.) and multimedia content. The data is stored in JSON format. An example is shown below: { "user_info": { "name": "Zhang San", "position": "Marketing Manager", "company": "XX Technology Co., Ltd." "skills": ["Marketing", "Digital Marketing", "Brand Planning"], "resources": ["30+ Partner Media"]}, "contact": "138xxxx1234", "services": ["Enterprise Brand Full Case Planning"], "requirements": ["Seeking brand promotion projects with an annual budget of more than 1 million"], "visual": { "style": "Simple business style", "image_url": "https: / / example.com / card_image_123.jpg"}} Preferably, in this embodiment, obtaining business card information input by a user and generating the user's first business card data using an artificial intelligence model includes: obtaining user identity information and / or business card style information and / or user capability information and / or user business resource information and / or user-provided business product / service information and / or user demand information input by the user; invoking a text-based artificial intelligence model to generate text for the user's first business card; invoking an image-based artificial intelligence model to generate an image for the user's first business card; and constructing the user's first business card data based on the text and image in the generated first business card. The following provides a detailed explanation using a practical example: Users enter business card information in the following ways: Structured form input: User identity information: including name (required), position, company name, contact information (phone / email), industry (select from drop-down menu, such as "Information Technology / Consumer Electronics / Education and Training"), and years of experience (numeric input); Business card style information: provides preset style options ("Simple Business Style," "Futuristic Technology Style," "Luxury High-end Style," "Fresh and Simple Style"), and supports users to upload custom style descriptions (e.g., "Must include the company's primary color blue, and use an asymmetrical layout"); Capability information: supports two input methods: Keyword tag selection (preset tag library: "Marketing Promotion", "Product R&D", "Supply Chain Management", "Cross-border E-commerce", etc., custom tag additions are available); Text description (e.g., "Experience in cold-starting SaaS (Software as a Service) products from scratch, and proficient in SEM / SEO optimization"); Business resource information: Enter in list form, for example: [ {"resource type":"cooperative media","details":"30+ vertical industry media (including 3 platforms with tens of millions of traffic)"}, {"Resource Type":"Channel Network","Details":"Offline dealer network in 20 cities nationwide"}, {"Resource Type":"Technology Patent","Details":"Owns 5 patents for data encryption technology"}] Product / Service Information: You can upload product manuals (PDF / PPT), service case links, or use a rich text editor to describe the service content (e.g., "We provide customized development of enterprise-level AI customer service systems, including natural language processing model (NLP) training, multi-channel access, and data visualization reports"); Demand Information: Set demand categories ("Seeking Cooperation," "Purchasing Requirements," "Talent Recruitment") and provide structured fields (for example, "Seeking Cooperation" requires the target partner type, expected cooperation scale, and core requirements).

[0026] File import and optical character recognition (OCR): This feature allows users to upload images of existing paper business cards and automatically recognize basic information such as name, phone number, and company name through OCR technology (such as the Baidu AI Open Platform OCR interface). Users can also manually modify the recognition results.

[0027] In this embodiment, the specific calling logic of the text-based artificial intelligence model is as follows to generate the text of the user's first business card: Model selection and input parameters: Basic text generation: Using the ChatGPT-4 model, the input parameters include: { "prompt": "Generate professional business card text based on the business information provided by the user. The style is {style description}. It must include core information: {name}, {position}, {company}, {core capabilities}, {products and services}, {contact information}. The language should be concise and professional, and the layout should be in point format." "temperature": 0.6, "max_tokens": 200} Industry-customized generation: For specialized industries such as finance and healthcare, we use domain-fine-tuning models (such as ChatGPT-4-Finance, which is trained based on financial corpus) to ensure the accuracy of professional terminology (for example, automatically adding keywords such as "compliance qualifications" and "risk control" in the financial industry).

[0028] Generate result processing: Perform structured analysis on the text content output by the model, extract key information fields (such as name, position, and service scope), and perform consistency verification with user input data (for example, checking whether the company name matches); Users are supported to manually edit the generated text (such as adjusting the layout, adding details), and the edited content is automatically synchronized to the structured data layer.

[0029] In this embodiment, the visualization generation logic of the image-based artificial intelligence model is as follows, generating an image of the user's first business card: Style parsing: Convert the user-selected style (e.g., "Technology and Future") into a prompt that the DALL-E model can recognize. Example: "A business card background with a futuristic technology style,featuring blue and purple gradient, circuit board patterns, and a minimalistlayout, suitable for a tech company executive." Element fusion: Based on the company logo uploaded by the user (supporting PNG / SVG formats), the logo is embedded into the generated background image through image synthesis technology (such as the Python Pillow library), and the position can be automatically adapted (top left corner / center / bottom right corner); Layout design: Adopting a responsive layout algorithm, the image-text ratio is automatically adjusted according to the length of the text content (such as a 1:2 or 1:3 image-text ratio), ensuring that the text is fully displayed and visually balanced.

[0030] Output specifications: The generated business card image resolution is 1000×600 pixels (suitable for electronic business card display), includes CMYK color mode (meets printing requirements), and provides a transparent background version (for digital scene embedding).

[0031] In this embodiment, the construction and storage logic of the first business card data is as follows: Data integration logic: Parse the generated text content into JSON structured data (including fields: name, position, company, capability label, service description, requirement type); Image data is stored as an object containing metadata, for example: { "image_metadata": { "url": "https: / / cdn.example.com / card_image_abc123.png", "style": "Technology and Future Style", "logo_url": "https: / / cdn.example.com / company_logo_xyz456.svg", "layout": "left-image-right-text"}, "text_content": "Zhang San\nChief Technology Officer\nXX Technology Co., Ltd.\n10 years of experience in AI algorithm R&D\nLed the launch of three products with tens of millions of users\nProvides customized development of intelligent customer service systems\nConsultation hotline: 138xxxx1234"} Use the data verification interface to ensure that the text and image information are consistent (for example, the company name matches the text and the logo).

[0032] Multimodal output supports: Generate electronic business cards (HTML5 format, supports click-to-dial and redirect to official website); Generate an API interface for third-party systems to call (such as the Customer Relationship Management System (CRM) to directly read structured capability information).

[0033] Preferably, the business card information input by the user is obtained, and video data for the user's first business card is further generated using an artificial intelligence model, forming multimodal first business card data, i.e., the first business card data includes text, images, and video. For example, based on the above-mentioned embodiment of generating text and image business cards, a multimodal large model (such as GPT-4V, CLIP, etc.) is introduced, combined with user dynamic knowledge base data, to construct a video business card generation framework that "real-time semantic understanding - scenario-based content generation - dynamic element fusion", realizing the following technical solutions: Knowledge base-driven video content generation: Utilizes unstructured data such as historical project cases, product update documents, and technology patents in the user's (individual / enterprise) knowledge base to generate dynamic video scripts through large-scale model training; Real-time update mechanism: When the knowledge base data changes (such as new cooperation cases, product parameter updates), the video content reconstruction is automatically triggered to ensure that the business card information is synchronized with business dynamics.

[0034] The specific implementation steps include the following technical solutions.

[0035] Knowledge base data extraction and preprocessing: Individual users: Extract project experience documents (such as PPTs and case reports), public speech video clips, industry sharing articles, etc. from their knowledge base, and use OCR and NLP to extract core information (such as "Leading the smartwatch promotion project in 2024 and achieving 100,000+ user conversions"); Enterprise users: Extract product manuals, technical white papers, and brand promotional materials (such as product demonstration videos and exhibition photos), and query core technology nodes in the enterprise knowledge graph (such as "flexible screen manufacturing process" and "blood oxygen monitoring patent") through the graph database.

[0036] Sample input (enterprise knowledge base snippet): { "product_updates": ["Smartwatch Pro version will be released in Q2 2025, adding heart rate warning function"], "case_studies": ["Completed new product pre-launch activities with XX Media, with exposure reaching over 5 million"], "tech_patents": ["Patent No. ZL202410012345, flexible screen bending life extended to 100,000 times"]} Multimodal large model generation video script and materials: Calling a large multimodal model (such as GPT-4V), the input parameters include: { "prompt": "Generate a video script for a tech company's new product promotion. Use a futuristic, business-themed theme. Include the company's core technologies (flexible screens, blood oxygen monitoring), product upgrades for Q2 2025, and successful collaboration examples. The video storyboard should incorporate a dynamic interpretation of the company's logo and data visualization charts.", "knowledge_source": "Enterprise Knowledge Base 2025 Q2 Update", "visual_style": "Reference to the technological blue-purple gradient background generated by DALL-E, with 3D product model rotation animation"} The model outputs structured scripts (such as storyboard descriptions, subtitle text, and animation effects), and calls a video generation model (such as Runway ML) to generate dynamic clips based on the scripts and corporate materials (logos, product images).

[0037] Real-time fusion of dynamic elements: Time trigger: Automatically scan the knowledge base on the 1st of each month. If a product update or new cooperation case is detected, regenerate the corresponding data visualization module in the video (such as replacing the case data chart); Event triggering: When a user participates in a new business event, an event tag (such as "2025 Global Smart Wearable Summit Exhibitor") is automatically added to the end of the video and linked to the official website of the event.

[0038] Multimodal business card integration: The generated video (MP4 format, resolution 1920×1080, supporting 15-30 seconds loop playback) is encapsulated with the original text and image data into a multimodal JSON structure.

[0039] Step 101: Obtain the business project demand data input by the user In this embodiment, users input business project demand data through the project demand entry module, which supports two input methods: Enter the project name (e.g., "2024 New Product Online Promotion Project"), project type ("Brand Promotion"), budget range (500,000-1.5 million RMB), execution period (3 months), and core objective ("Increase new product awareness and achieve 100,000+ potential customer conversions"). Natural language description input: Users directly enter "need to plan an online promotional campaign for a newly released smartwatch, focusing on white-collar workers aged 25-35, with a budget of less than 1 million yuan and a three-month duration." The system then uses an NLP model (such as BERT, Bidirectional Encoder Representations from Transformers) to parse this into structured data. Ultimately, this data is integrated into standardized business project requirements data, including fields such as basic project information, target group, budget, duration, and core requirements, and stored in a database for subsequent processing.

[0040] Step 102: Generate business project solution data based on the business project demand data In this embodiment, preferably, a preset business project implementation model (a deep learning model trained based on historical successful project data) is called to process the demand data obtained in step 101: Demand analysis: Extract core elements, such as "new smartwatch product promotion," "white-collar workers aged 25-35," "1 million yuan budget," and "3-month cycle." Sub-step generation: Break the project down into three core sub-steps: Sub-step A: Market research (2 weeks, objective: to identify target group’s media habits and competitive product trends); Sub-step B: Plan Design (4-week cycle, goal: develop a combination plan for social media marketing, KOL (Key Opinion Leader) collaboration, and information flow advertising); Sub-step C: Execution and optimization (cycle: 6 weeks, goal: implement the promotion according to the plan and adjust the strategy in real time).

[0041] Resource Matching: Based on the user capability information ("Digital Marketing Expertise") and business resource information ("30+ Partner Media") in the first business card data generated in step 100, corresponding second business card data candidates are generated for sub-step B (e.g., "Technology Blogger @Digital Xiaolingtong" in the KOL resource library, "XX Media, Toutiao Advertising Agent" in the media resource library). Preliminary business partnership plan data is also generated (e.g., "KOL partnership budget: 300,000 yuan, plans to publish 3 in-depth review videos"). The final output is a business project plan data, including sub-step breakdown, resource matching recommendations, and timeline planning, presented as a structured document.

[0042] Step 103: Match at least one second business card data with the business activity artificial intelligence model according to the business project plan data to form a second business card data set corresponding to the user's business project needs. In this embodiment, a business activity AI model is trained to match the business cards of business partners involved in the business project plan based on the business project plan data, thereby meeting the user's business project needs. This business activity AI model can parse and recognize the text in the business project plan and efficiently match partners that can meet the user's business project needs (i.e., the business partner corresponding to the second business card).

[0043] Preferably, in this embodiment, the business activity artificial intelligence model adopts a multi-level processing architecture: Semantic Parsing: Use the BERT model to perform semantic parsing on business project plan data and generate a structured semantic architecture. For example, "smart watch new product promotion" can be parsed into {"field":"consumer electronics","product type":"smart wearable devices","project type":"new product promotion"}; Entity extraction: Extract domain entities from the semantic architecture, including "smartwatch," "white-collar workers aged 25-35," "information feed advertising," and "KOL collaboration." Knowledge Graph Retrieval: Based on the above entities, query conditions are constructed, and a candidate set of second business card data that meets the conditions is retrieved from the business activity knowledge graph (data sources include: business card data (more than 100,000 enterprise and personal business cards); enterprise industrial and commercial information (connected to Tianyancha API, covering more than 80 million enterprises); product service description information (integrating industry reports and enterprise official website data)). For example, for the "KOL cooperation" entity, 3 KOL business cards (@Digital Lingtong, @Tech Big Cousin, @Smart Life Expert) with a fan base of more than 100,000 and a history of cooperation with smart wearable devices are retrieved. For the "information flow advertising" entity, 5 business cards of advertising agents with experience in the consumer electronics industry are retrieved. Finally, the top 5 second business card data with high matching degrees are selected through the cosine similarity algorithm (threshold set to 0.75) to form a set of 10 candidate second business card data.

[0044] Preferably, for the construction of the business activity artificial intelligence model, a federated knowledge graph architecture is further added.

[0045] For example, the following federated knowledge graph architecture design: Participants: Core enterprise A (user), upstream supplier B, downstream channel partner C, each holding a local knowledge base (such as enterprise A's technology patent library, enterprise B's raw material production capacity data, enterprise C's channel coverage data); Federated learning mode: Vertical federated learning is adopted. Under the same sample ID space, each enterprise collaboratively trains a "demand - resource" matching model. The feature space includes: Enterprise A: Project requirements (budget, technical requirements), its own ability labels; Enterprise B: Raw material supply capacity (production capacity, certification qualifications), historical cooperation cases; Enterprise C: Channel coverage range (region, customer group), promotion case effects.

[0046] Cross - enterprise data collaborative training Phase 1: Local feature extraction Each enterprise uses local data to train a sub - model. For example, enterprise A uses BERT to extract demand text features, and enterprise B uses a CNN (Convolutional Neural Network) model to extract product parameter features, generating encrypted intermediate feature vectors; Phase 2: Federated parameter aggregation Through the secure multi - party computation (MPC Secure Multi - Party Computation) protocol, without exposing the original features, each enterprise sends the sub - model parameters to a trusted aggregation node to generate a global "demand - resource" matching model (such as a federated neural network). Example aggregation algorithm: # FedAvg simplified logic def federated_aggregation(local_models): weights = [model.weight * len(model.data) for model in local_models]return sum(weights) / sum(len(model.data) for model in local_models) Phase 3: After the global knowledge graph is updated and trained, each enterprise receives the global model parameters and updates the relationship weights of the local knowledge graph (such as the matching score of "Enterprise A-Demand-Enterprise B-Supply Capacity").

[0047] Federated reasoning and collaborative solution generation When a user (Company A) enters a project requirement (e.g., "We need a flexible screen supplier with a monthly production capacity of 100,000 units for mass production of the Pro version of the smartwatch"), the system processes it through the following steps: Demand analysis: Use the global model to extract feature vectors (e.g., "flexible screen" and "monthly production capacity of 100,000"); Federated reasoning: Each enterprise's local model calculates the matching degree based on the encrypted feature vector and returns an encrypted list of candidate suppliers (for example, Enterprise B's matching score is 0.85). Privacy protection solution generation: Using zero-knowledge proof (ZKP) technology, company B proves to company A that it has a monthly production capacity of 100,000+ units without revealing specific production capacity data. Example proof logic: { "proof": "Verify through zero-knowledge proof that company B has the ability to produce ≥ 100,000 flexible screens per month", "encrypted_data": "Enterprise B's production capacity certificate ciphertext", "verification_key": "Public key used to decrypt the verification"} The resulting collaboration plan includes cross-enterprise collaboration recommendations, such as: { "collaboration_suggestion": [ { "partner": "Company B (flexible screen supplier)", "federated_score": 0.85, "proof_of_capability": "ZKP Proof Document", "joint_plan": "The capacity matching plan recommended by federated learning: Company A provides technical parameters, Company B is responsible for production, and the cycle is 45 days."} ]} Preferably, in this embodiment, data is further collected from business card data and / or third-party business information and / or product and service descriptions and / or user social content on social platforms and / or commercial contract transaction information in electronic commercial contracts signed during commercial activities; commercial activity knowledge is then mixed and extracted from this information; and the commercial activity knowledge of upstream and downstream commercial activity entities in the user's commercial activity industry chain is integrated to form a global commercial activity knowledge graph. This global commercial activity knowledge graph is then used to match the commercial project plan data to retrieve a second set of qualified business card data candidates.

[0048] For example, the construction of a knowledge graph of business activities is preferably carried out in the following manner. The multi-source data acquisition system collects the following data in real time through the data interface: Business card data: The business / personal business card of the user who registered the business card in this embodiment, including structured data such as the company name (such as "Shenzhen Zhilian Technology Co., Ltd."), main business (smart wearable device research and development), core technology (flexible screen manufacturing process), and contact information.

[0049] Business registration information: Obtain business registration information such as registered capital (50 million yuan), business scope (wearable device sales), and equity structure (major shareholder A holds 35% of the shares) through the National Enterprise Credit Information Publicity System. Product service description: Crawl unstructured text from company official website product detail pages, such as "Zhilian Technology" smartwatch product parameters (screen size 1.39 inches, battery life 14 days), technical advantages (blood oxygen monitoring accuracy ±1%), etc. User social content: Capture technical sharing posts by companies in industry forums, such as the "Smartwatch Battery Life Optimization Solution" posted by an engineer, which contains technical keywords such as lithium-ion battery model and energy density.

[0050] Commercial contract transaction information: The processing contract between Zhilian Technology and the OEM factory is extracted from the historical cooperation database in this embodiment, including transaction terms such as delivery cycle (45 days) and quality standards (ISO9001 certification). Then the hybrid knowledge extraction process is performed. Structured data processing: Use SQL statements of relational databases to directly extract the basic attributes of enterprises in industrial and commercial information, and construct triples such as "Enterprise - Registered capital - 50 million". Unstructured text extraction: Use the BERT-NER model (Bidirectional Encoder Representations from Transformers for Named Entity Recognition, a named entity recognition model based on bidirectional encoder representations) to perform named entity recognition on product descriptions and social content, and extract domain entities such as "Smartwatch", "Flexible screen", "Blood oxygen monitoring", as well as relationship types such as "Manufacturing process", "Optimization plan". Semi-structured data parsing: Parse contract terms through regular expressions to extract key information such as "Party A - Zhilian Technology", "Party B - Dongguan Huayu Electronics", "Delivery cycle - 45 days". Knowledge fusion and graph construction Use entity disambiguation technology (such as name matching based on cosine similarity) to merge information from different data sources of the same entity. For example, associate the industrial and commercial information, product data, and social content of "Zhilian Technology" into a unified node. Construct a global business activity knowledge graph through the Neo4j graph database. The node types include: Enterprise entities (Zhilian Technology, Huayu Electronics) Product services (Smartwatch manufacturing, Flexible screen manufacturing) Technical capabilities (Blood oxygen monitoring, Supply chain management) Business relationships (Cooperation relationships, Upstream and downstream relationships) The edge types are defined as "Provide products", "Possess technology", "Have cooperated", etc., forming an industry knowledge network containing more than 100,000 nodes and more than 500,000 relationships.

[0051] Step 104: Establish business information processing between the first business card data and the second business card data based on the second business card data set In this embodiment, a business information processing list can be generated based on the second business card data set. This list specifically includes: Data association: Establish a two-way association between the first business card data (User Zhang San) and the second business card data (such as KOL@Digital PHS), and mark the association type ("Project cooperation candidate") and the association strength (based on the matching score, range 0 - 100, here it is 85 points); Interaction interface: Provide three information processing methods: Solution sharing: Push commercial project solution data (encrypted PDF format) to the candidate object, along with a statement of cooperation intention (such as "We sincerely invite you to participate in the new product promotion project of smartwatches and look forward to communicating the cooperation details"); Demand matching: Open online communication channels (built-in IM instant messaging tool) to support users and candidates to exchange project details in real time; Process start: After the user confirms the cooperation, a project collaboration task is automatically created (for example, a "KOL cooperation signing" task is generated in the project management system and the person in charge is assigned to Zhang San).

[0052] Example 2: In order to more clearly reveal the working details of the commercial information processing system of the present invention, the framework principle diagram of the commercial information processing system provided in the second embodiment of the present invention is shown in FIG. Figure 2 .

[0053] The commercial information processing system 200 provided in the second embodiment of the present invention includes: a client 201, a cloud development layer 202, a third-party service 203, and an AI service layer 204. The operation of the commercial information processing system 200 includes the following process: 1. AI Business Card Creation Process Client 201 initiates a business card creation request via WeChat mini-program 2011. AI business card module 2012 collects basic business card information and style preferences, which are then transmitted to cloud development layer 202 via an encrypted channel. Basic business card information includes user capabilities, business resources, product / service offerings, and / or user needs. This information can be stored in cloud database 2021.

[0054] Parallel execution of cloud AI processing: Cloud function 2024 on the cloud development layer 202 calls AI model 2041, which deploys the ChatGPT-4 model to process text (for example, rewriting "chip sales" to "high-precision semiconductor solutions"). Cloud function 2024 then calls image generation API 2042 to create a business card background (prompt example: "Technology blue minimalist style, containing quantum computing elements"). Result of receiving and processing generated content: The text processed by the AI big model 2041 is stored in the cloud database 2021; The image generated by the image generation API 2042 is stored in the cloud database 2021; The processing result is returned to the WeChat applet 2011 of the client 201, and the business card preview page is dynamically rendered (WebGL realizes 3D flipping special effects) to complete the local cache.

[0055] 2. Connection Recommendation Process The client 201 collects user behavior data (clicks / stays / shares) in real time through the social circle module 2013 in the WeChat applet 2011, constructs a spatiotemporal feature vector (256 dimensions), and pushes it to the cloud development layer 202.

[0056] The social circle module launched dual-engine retrieval in 2013: the Neo4j2023 graph database performs path queries (for example, finding AI experts in second-degree connections) and Elasticsearch2022 performs semantic search (BM25+BERT dual-channel matching).

[0057] The AI service layer 204 runs a hybrid recommendation model, integrates the dual-engine search results, integrates the analysis results of the social network analysis 2043, injects business rules (such as prioritizing local resources), generates a top 50 candidate list, and feeds it back to the network circle module 2013 for loading and display. The recommended top 50 candidate list is pushed through the WeChat interface, and the sharing card carries a dynamic signature (HMAC-SHA256 anti-tampering).

[0058] 3. Business Cooperation Process The client 201 obtains structured business project demand data (for example, 7nm chips are required, with a monthly supply of 100,000 pieces) in real time through the business cooperation module 2014 in the WeChat applet 2011, triggering the cloud function 2024 in the cloud development layer 202 for intelligent analysis.

[0059] The cloud development layer 202 performs multi-dimensional matching: Neo4j query supplier qualification path (example: TS16949 certified and serving car companies) Elasticsearch (Elasticsearch) retrieval history collaboration case (TF-IDF weight optimization, i.e., Term Frequency-Inverse Document Frequency weight optimization) The AI service layer 204 generates business project plan data through the AI big model 2041: for example, using ChatGPT-4 to write technical terms (citing industry standards in the knowledge base); for example, using the GNN model (Graph Neural Network) to predict the success rate of cooperation (inputting the supply chain relationship map).

[0060] Cloud function 2024 calls third-party service 203: generates legal documents and contracts through electronic contract API 2031 (automatically fills in company information) for signing; initiates the payment process by calling the WeChat payment interface (supports staged payment) to realize the payment link in the contract document.

[0061] The third-party service 203 synchronizes the performance data in real time to the cloud database 2021 of the cloud development layer 202 for storage, updates the supplier rating and production capacity data, and analyzes and processes the updated data in the social network analysis 2043 and the AI big model 2041 to form a closed loop.

[0062] Preferably, in this embodiment, at least one third business card data is matched based on the user's capability information and / or user's business resource information and / or business product / service information provided by the user and / or user demand information in the first business card data stored in the cloud database 2021 to form a third business card data set corresponding to the first business card data; and based on the third business card data set, business information processing is established between the first business card data and the third business card data.

[0063] In this preferred embodiment, the information in the first business card data can be directly matched to obtain the corresponding third business card data, thereby forming a communication and processing of business information. Figure 2 The modules in the commercial information processing system 200 are disclosed in detail.

[0064] First, multi-dimensional information extraction is performed. The specific process of obtaining information is as follows: The first business card data is retrieved from the AI Business Card Module 2012 and the Cloud Database 2021, and core dimension information is parsed from it. Capability Information: Extracts user-annotated professional skill tags (such as "blockchain development" and "international commercial arbitration"), professional qualification certifications (such as PMP project management certification, CPA qualification), and past project experience scores. Business Resource Information: Parses a list of sharable resources, including offline channels (such as "30+ dealer network in North China"), technology platforms (such as "independently developed AI customer service system API interface"), and capital reserves (such as "available angel investment quota of 5 million yuan"). Product / Service Information: Structures user-provided product parameters (such as "industrial-grade 3D printer printing accuracy ±0.1mm"), service processes (such as "cross-border e-commerce full-link operation service including 8 standard nodes"), and after-sales guarantee terms. Demand Information: Identifies user-specified cooperation requests, such as "Seeking new energy vehicle battery production contract manufacturing" and "Recruiting city partners in East China."

[0065] The business information processing system 200 establishes a multi-source information fusion mechanism through third-party services 203. This mechanism simultaneously accesses historical user behavior data within the business information processing system 200 (e.g., in-depth browsing history of 50+ supply chain companies over the past six months), business social graphs (automatically importing 100+ verified contacts from third-party platforms like LinkedIn), and third-party credit data (authorized access to corporate credit ratings and legal dispute records). Natural language processing technology is used to extract entities from unstructured information (e.g., user-posted letters of intent for cooperation and project bid summaries), generating user profile vectors containing over 200 features.

[0066] In the AI service layer 204, the AI model 2041 constructs a hybrid recommendation model based on the attention mechanism. First, feature fusion is performed: the user profile vector is aligned with the enterprise / individual feature vectors (covering over 150 dimensions, including core business, resource advantages, and cooperation history) in the existing business card database in the cloud database 2021. Similarity calculation is then performed: the cosine similarity algorithm is used to calculate the matching degree, focusing on capability complementarity (40%), resource synergy (30%), and demand alignment (30%). Finally, a threshold filter is performed: a filter condition of matching degree ≥ 0.7 is set to generate a third business card data set containing at least three highly correlated objects. For example, a user providing "cross-border e-commerce independent website construction" services may obtain a precise set of "overseas social media marketing agencies," "cross-border payment solution providers," and "international logistics service providers" after matching.

[0067] Next, in this preferred embodiment, the generated third business card data set can be visually displayed. In the client 201 interface, such as the business cooperation module 2014, the core information of the third business card (subject name, matching tags, and historical interaction records) is presented in the form of a node diagram. Node connections are annotated with capability complementarity points (e.g., "Technology Output -> Resource Input") and demand matching status ("pending communication," "preliminary negotiation," and "cooperation"). Intelligent information synchronization is implemented. When the third business card subject updates key information (e.g., newly added patented technologies, major cooperation developments), the system uses natural language processing (NLP) technology to extract core elements and generate customized push copy (e.g., "XX supply chain company you're interested in has launched a new Southeast Asian shipping line, which is highly compatible with your cross-border e-commerce business"). Secure interaction mechanism: Preferably, all business information interactions are recorded and stored via blockchain, for example, preferably in the cloud database 2021 by accessing the blockchain. Sensitive data transmission utilizes the national SM4 encryption algorithm, and users can customize the scope of information sharing in privacy settings (e.g., only opening core technology parameters to the top three matching partners).

[0068] Preferably, in this embodiment, the matching of at least one business card data includes: obtaining business card data, user business friend information, user behavior data, and user chat content data; analyzing and generating a social relationship graph according to a business social network analysis model; generating a user interest vector according to a user behavior model; and generating at least one matching business card data in a hybrid recommendation model based on the social relationship graph and the user interest vector.

[0069] The details of the solution for matching at least one business card data in the preferred embodiment are as follows: Obtain the first business card data stored in the cloud database 2021, such as the user's ability information ("with 5 years of marketing experience"), business resource information ("30+ cooperative media"), and user demand information ("seeking brand promotion projects with an annual budget of over 1 million"), or the business product / service information provided by the user's own company ("Internet traffic advertising service or advertising placement platform product"); Further initiate the following data acquisition for a multi-dimensional matching mechanism: The cloud database 2021 synchronously collects user business friend information (for example, Zhang San's friend list includes 10 advertising companies and 5 KOLs), user behavior data (viewed 15 intelligent wearable industry reports and downloaded 3 competitor analysis documents in the past 30 days), and user chat content data (mentioned "channels for precise targeting of white-collar groups" many times in conversations with colleagues); Social relationship analysis: Construct a social relationship graph through the business social network analysis model (using the GNN graph neural network) in the social network analysis 2043, and identify that Zhang San has a second-degree personal connection with Advertising Company A (the common friend is a media director) and has a direct cooperation history with KOL@Digital Smartphone; Interest vector generation: Use the user behavior model (TF-IDF algorithm) in the AI large model 2041 to convert the browsing records, downloaded content, etc. into interest vectors, with dimensions including {"Smart Wearables": 0.8, "White-Collar Marketing": 0.7, "In-Feed Advertising": 0.6}; Hybrid recommendation: Input the social relationship graph and interest vector into the hybrid recommendation model (combining collaborative filtering and content recommendation algorithms) in the AI large model 2041 to generate the matching third business card data, such as recommending "Vertical Media 'Weekly for Workplace Elites' Focusing on White-Collar Groups" (matching score: 88 points), "Hangzhou Convention and Exhibition Center, the Organizer of the Intelligent Wearable Industry Exhibition" (matching score: 85 points), and forming a set of 8 candidate third business card data.

[0070] Business information processing expansion: A deep association is established between the third business card data set and the first business card data, allowing for recommended display within WeChat mini-program 2011. For example, AI model 2041 can preferably generate a customized partnership proposal for "Career Elite Weekly" ("Insert new product advertising in the magazine's special section, budget 200,000 yuan") and simultaneously update the user's business resource library stored in cloud database 2021, forming a dynamic business information network.

[0071] Example 3: To more clearly reveal the details of the business information processing system provided in the second embodiment of the present invention in generating a business project plan for processing, see Figure 3 , Figure 3 A schematic diagram of a process for generating and processing a business project plan by a business information processing system provided in a third embodiment of the present invention is provided.

[0072] See also Figure 3 In the third embodiment of the present invention, the business information processing system 200 provided in the second embodiment will be used to implement the entire process of generating and processing a cross-border e-commerce project plan. In this embodiment, the WeChat mini-program 2011 on the user side supports user requirements input, plan review, contract signing, and payment operations. Cloud Function 2024 is responsible for data preprocessing, matching algorithm invocation, and plan generation logic scheduling. AI Service 204 integrates the GPT-4 model to generate business plan sub-steps and cooperation strategies. AI Big Model 2041 (Knowledge Graph) stores industry knowledge (such as cross-border e-commerce policies, logistics provider qualifications, payment service provider rates), success cases, and corporate credit data. Third-Party Service 203 provides interfaces for electronic contract signing (such as FaDaDa), online payment (such as WeChat Pay), and credit inquiry (such as Qichacha).

[0073] The following example describes a project to build a Southeast Asian cross-border e-commerce platform. In WeChat Mini Program 2011, the user fills in a business project requirement description (business capabilities / products and services). For example, in this example, the requirements are: "Build a Southeast Asian cross-border e-commerce platform targeting the Thai and Vietnamese markets. Thai / Vietnamese localization is required. Lazada logistics integration (first-mile collection + international transportation + last-mile delivery) is required. GrabPay payment access is available (supporting THB / VND currencies). An MVP (Minimum Viable Product Launch) must be launched within three months. The budget is 800,000 yuan. Priority is given to technical teams, logistics providers, and payment service providers with Southeast Asian project experience." After submission, the Mini Program sends the business project requirement description to Cloud Function 2024.

[0074] Cloud Function 2024 performs data preprocessing (word segmentation and vectorization): Use the BERT model to segment the demand description text, extract the core entities ([Southeast Asia, cross-border e-commerce platform, Thailand, Vietnam, Thai, Vietnamese, Lazada Logistics, GrabPay, 3 months, MVP, 800,000, Southeast Asian project experience, technical team, logistics provider, payment service provider]), convert them into 1024-dimensional word vectors, and store them in the demand feature library.

[0075] The AI big model 204 can perform knowledge graph query and return the following related data: Industry knowledge: Southeast Asia's e-commerce market size in 2024 (US$32 billion in Thailand, US$28 billion in Vietnam), Lazada logistics rates (US$0.5 per order for the first kilometer, US$1.2 per order for the international leg, and US$0.3 per order for the last mile), and GrabPay rates (2.5% of the transaction amount).

[0076] Success story: "A brand's Southeast Asian platform construction case in 2023 (launched in 4 months, cost 750,000, technical team A + logistics provider B + payment provider C)".

[0077] Corporate credit investigation: Screen out technical teams with "Southeast Asian project experience ≥ 2 years" (such as TechTeam X, credit rating AA), Lazada certified logistics providers (such as LogiCorp Y, covering 70% of Thailand), and GrabPay authorized payment providers (such as PayPro Z, fee rate 2.3%).

[0078] Cloud Function 2024 returns the associated data through the following matching algorithm and then sorts it: This includes the calculation of cosine similarity: the similarity between the demand vector and the feature vector of the existing business card library (capability information of technical teams, logistics providers, payment providers, etc. and / or user business resource information and / or business product / service information provided by users) (such as "10 years of project experience", "Southeast Asian experience", "Lazada certification", "GrabPay authorization", "Amazon official service provider", "Southeast Asian government e-service procurement supplier") is calculated: Technical Team: TechTeam X (0.95), DevGroup Y (0.88), CodeLab Z (0.82) Logistics providers: LogiCorp Y (0.92), ShipCo Z (0.85), Transit X (0.80) Payment providers: PayPro Z (0.90), PayLink Y (0.85), PayServ X (0.80) Then perform collaborative filtering optimization: Combine historical user data (if available, it is better to use industry average data) and exclude the following businesses: Technical Team: DevGroup Y (Project delay rate 15% in the past 6 months, risk marker) Logistics provider: Transit X (20% complaint rate for final delivery, risk flagged) Payment provider: PayServ X (funds arrival delay ≥ 3 days, risk flag) Finally, the matching ranking results are generated based on the matching degree as follows (the business card data of these businesses can be provided to the client for corresponding display): Technical Team: TechTeam X (Priority 1), CodeLab Z (Priority 2) Logistics providers: LogiCorp Y (Priority 1), ShipCo Z (Priority 2) Payment providers: PayPro Z (Priority 1), PayLink Y (Priority 2) Next, cloud function 2024 requests AI service layer 204 to generate a business cooperation plan. The specific steps are as follows: GPT-4 calls and business cooperation plan generation are performed in the AI service layer 204. The specific steps are disclosed as follows.

[0079] Input the demand feature vector, the matching results (including the matched business entity's capability information and / or business resource information and / or provided business product / service information), and the knowledge graph data in the AI big model 2041 (cost calculation, time nodes, general contract templates, etc. in market dynamics) to generate the following solution sub-steps: For example, the sub-steps of the plan (3-month MVP launch plan): Total cost estimation: 5+300,000+3+2+5=450,000 yuan (the remaining 350,000 yuan is used for operating reserves, which meets the budget requirements).

[0080] Week 1-2: Demand Research and Architecture Design (matching TechTeam X, you can provide TechTeam X's business card data here) Output: Localization requirements document (Thai / Vietnamese interface specifications), Lazada logistics API integration solution, GrabPay payment integration solution.

[0081] Cost: Technical consulting fee of RMB 50,000 (TechTeam X, calculated based on man-days: 5 people × 2 weeks × RMB 500 / man-day).

[0082] Generate the corresponding commercial electronic contract data: "Technical Consulting Service Agreement" (automatically generated, including service content, deliverables, and acceptance criteria).

[0083] Week 3-6: System Development (matching TechTeam X + CodeLab Z, business card data for TechTeam X and CodeLab Z can be provided here) Tasks: Front-end interface development (TechTeam X, Thai / Vietnamese), back-end system construction (CodeLab Z, logistics / payment interface integration).

[0084] Cost: Development costs of 300,000 yuan (TechTeam X: 200,000 yuan, CodeLab Z: 100,000 yuan, calculated based on man-months: 8 people × 1.5 months × 2,500 yuan / man-month).

[0085] Generate the corresponding commercial electronic contract data: "Software Development Service Agreement" (acceptance in stages: delivery of front-end prototype in the 4th week and delivery of back-end interface in the 6th week).

[0086] Week 7-8: Logistics Integration (matching LogiCorp Y, you can provide LogiCorp Y business card data here) Task: Test the integration of Lazada’s logistics API (first mile, international segment, and last mile), and generate a logistics cost simulation model (calculated by order volume).

[0087] Cost: Interface integration fee of RMB 30,000 (LogiCorp Y, one-time fee).

[0088] Generate corresponding commercial electronic contract data: "Logistics Service Cooperation Agreement" (including rates, service scope, and exception handling terms).

[0089] Week 9-10: Payment Integration (matching PayPro Z, you can provide PayPro Z business card data here) Task: Integrate GrabPay payment channel, test multi-currency settlement (THB / VND), and generate payment rate simulation table.

[0090] Cost: Interface integration fee of RMB 20,000 (PayPro Z, one-time fee), subsequent transaction fee rate of 2.3% (deducted from platform turnover).

[0091] Generate the corresponding commercial electronic contract data: "Payment Service Cooperation Agreement" (including rates, fund arrival cycle, and risk reserve terms).

[0092] Week 11-12: Joint debugging and launch (full team collaboration) Tasks: System integration (technical team + logistics provider + payment provider), stress testing (simulating 1,000 orders / hour traffic), MVP launch (open registration in Thailand and Vietnam markets).

[0093] Cost: Joint debugging fee of RMB 50,000 (calculated on a man-day basis: 10 man-days × 2 weeks × RMB 250 / man-day).

[0094] Generate corresponding commercial electronic contract data: No new contract is added, and execution is based on the acceptance terms of the previous agreement.

[0095] In each of the aforementioned business plan generation sub-steps, a contract master model (preferably AI master model 2041, which generates contract data) generates commercial contract data based on the business project requirement data and the business project plan data (i.e., the business project plan data in the business plan sub-step). Preferably, AI master model 2041, serving as the contract master model, can also further invoke a commercial contract generation master model within third-party service 203 to generate the commercial contract data implemented within AI master model 2041. This will not be further detailed here.

[0096] Preferably, further verification and optimization of the business plan's sub-steps can be performed: Cloud Function 2024 transmits the plan back to WeChat Mini Program 2011, where users can view a Gantt chart showing the timelines, responsible individuals, deliverables, and other contractual data for each business plan sub-step. Parameters can be adjusted in WeChat Mini Program 2011. For example, if a user requests "shorten the development cycle to 2.5 months," ChatGPT-4 will recalculate: adding technical team manpower (CodeLab Z adds two people), increasing costs by 50,000 yuan (to a total of 500,000 yuan, still within budget), and updating the plan to "development in weeks 3-7, joint debugging in week 8."

[0097] Preferably, the cloud function 2024 calls the electronic contract API of the third-party service 203 to generate the contract generated for each sub-step into a contract electronic signature link that can be used for electronic contract signing, so that both parties can sign and pay for execution.

[0098] The user and the partner receive the signing link through WeChat Mini Program 2011 and complete the electronic signature (preferably using blockchain for evidence storage and hash value storage to ensure tamper-proof). The user's business card data and the partner's business card data are loaded and displayed in WeChat Mini Program 2011 and automatically filled into the electronic contract corresponding to the received electronic contract signing link, enabling the user and the partner to edit, modify, and electronically sign after confirmation.

[0099] Next, the online payment and status tracking process is carried out through the payment interface of the third-party service 203. In this embodiment, during weeks 1-2: 50% of the consulting fee to TechTeam X (25,000 RMB, paid via WeChat Pay) is prepaid. During weeks 3-6: 50% of the development fee is prepaid (150,000 RMB, with the remaining 50% paid after staged acceptance). Subsequent costs include logistics / payment integration fees (30,000 RMB + 20,000 RMB, paid in one lump sum) and joint debugging fees (50,000 RMB, paid after launch).

[0100] Transaction status is also updated simultaneously. For example, Cloud Function 2024 will synchronize in real time: "TechTeam X has signed, payment of 25,000 yuan has been successful, and demand research has begun." The mini-program interface displays the status of each sub-step (such as "Development (30% complete)" and "Logistics integration awaiting signature"), and click to view detailed progress reports (such as code submission records and test reports).

[0101] Preferably, the business cooperation status from the business plan sub-step to the next sub-step is further updated, such as exception handling and iterative optimization during the execution of the sub-step.

[0102] For example, if LogiCorp Y suddenly cancels its Lazada certification (the knowledge graph is updated in real time), a risk alert will trigger Cloud Function 2024: the matching algorithm will be rerun, screening ShipCo Z (priority 2, valid Lazada certification). The GPT-4-generated solution will be adjusted: in weeks 7-8, the logistics integration will be replaced with ShipCo Z, increasing costs by 10,000 yuan (40,000 yuan for the interface integration fee), and the contract will be automatically updated. The user will receive a notification: "Logistics provider changed, solution optimized, new contract pending signing." After confirmation, the process will continue.

[0103] The business collaboration status data from the business plan sub-step to the next sub-step is fed back to the AI model 2041 for iteration. For example, after a project is completed, user reviews (e.g., "TechTeam X's development quality is high, and they delivered one week ahead of schedule") are stored in the knowledge graph and used to optimize subsequent matching algorithms (e.g., increasing TechTeam X's matching weight). Cost data (e.g., actual 450,000 yuan vs. budgeted 800,000 yuan) is fed back to the AI model 2041 to train the model for more accurate cost estimation capabilities.

[0104] Example 4: To more clearly reveal the details of the business information processing performed by the enterprise user in the business information processing system 200 provided in the second embodiment of the present invention, see Figure 4 , Figure 4A schematic diagram of a process for business information processing by an enterprise user provided in a fourth embodiment of the present invention is provided.

[0105] See also Figure 4 In this fourth embodiment of the present invention, an example of commercial information processing within a commercial information processing system 200 is presented, using an intellectual property service platform enterprise as an example. This intellectual property service platform enterprise's core businesses include patent application agency services, trademark registration, copyright registration, and intellectual property strategic consulting. Its service advantages include over 300 professional agents (including 20 experts in AI / IoT fields), and the provision of free patent search and technical solution evaluation. The enterprise's user administrator enters the company's business card information in the WeChat Mini Program 2011, including: Qualifications: National Intellectual Property Administration filing agency (number: XXX), ISO9001 certification (certificate number: XXX). The service system includes the patent application process (search-->writing-->submission-->examination-->authorization), agent grading (categorized by technical field and years of experience, such as "Gold Medal Agent in the AI Field"). The case library stores over 1,000 successful cases (such as "Full-Process Patent Application for a Certain AI Algorithm"). The mini program then encrypts the business card information and sends it to the AI Business Card Module 2012. The AI Business Card Module 2012 then requests the cloud function 224 to generate the business card data. For example, based on the input business card information, the mini program generates the business card text, "XX Intellectual Property, specializing in patent agency in the AI / technology fields, with a team of over 300 experts, safeguarding innovation," using the text AI model (ChatGPT4) in the AI Service Layer 204. It also generates a visual prompt, "Technology blue as the primary color, embedded with elements such as patent certificates and gears (symbolizing technological innovation)." This is then output to, for example, the DALL-E2 image generation interface to generate the image required for the business card. The generated business card text and image are then returned to the AI Business Card Module 2012 and stored in the cloud database 2021 as the business card generated for the enterprise user. Furthermore, the business card data of the enterprise user can be linked with the business card data of at least one employee of the enterprise. For example, a practicing patent agent within the enterprise can associate their personal business card data with the business card data of the enterprise user and store it in the cloud database 2021.

[0106] At the same time, enterprise users trigger the construction of an enterprise knowledge base (including patent law provisions, examination guidelines, agent skill matrix, fee standards, etc.). The enterprise's private domain data is then trained and synchronized with the knowledge base. For example, enterprise service processes, agent information, and industry rules (such as the 2025 invention patent official fee adjustment) are imported into the AI Big Model 2041 (the intellectual property service knowledge base model), forming a knowledge graph covering "patent application - agent matching - fee assessment - layout strategy."

[0107] After the AI model 2041 (the enterprise product and service knowledge base model) is trained, it can receive responses to individual users' product and service requests. For example, user Li Si submits a request for a patent application for an AI-powered garbage sorting system (including process, agent, fees, and layout) through WeChat Mini Program 2011. He uploads a technical proposal and a competitive product report. The Mini Program receives the request, pre-processes it (segmentation / vectorization), and sends it to Cloud Function 2024.

[0108] Cloud Function 2024 pre-processes product and service demand data and performs word segmentation / vectorization: for example, extracts keywords such as "AI garbage classification, invention patents", and analyzes technical solutions (intelligent algorithms and hardware terminals are high-value features) in order to match them with the business cooperation business card data in the cloud database 2021, such as matching the business card data of the practicing agents under the enterprise, thereby matching AI gold medal agents.

[0109] Furthermore, the AI big model 2041 (enterprise product service knowledge base model) infers and outputs the sub-steps of the business plan based on the product service demand data: for example Sub-step 1: Secondary patent search (3 days, output of patent application risk report); Sub-step 2: Matching a patent agent (matched with AI gold medal agent, Mr. Wang); Sub-step 3: Fee calculation: official fees (900 + 2500 yuan), agency fees (8000 yuan, including corrections); Sub-step 4: Patent application layout (domestic dual patent application strategy, overseas PCT).

[0110] Cloud function 2024 requests AI service layer 204, calls AI big model 2041 (enterprise product service knowledge base model), and generates a complete business cooperation plan based on business cases / industry rules (such as patent examination standards).

[0111] The final output of the business cooperation plan and matching business card data is as follows: Complete business cooperation plan: process manual (pictured steps + precautions), agent profile (Mr. Wang's resume + case studies), fee details (invention + utility model agency fees / official fees), layout strategy (SWOT analysis + domestic and international layout).

[0112] Business card data matching: Cloud database 2021 matches agent business card data (contact information, schedule), service nodes (search agency, payment channel), and returns the complete solution to WeChat mini program 2011.

[0113] Embodiment 5: In order to more clearly reveal the working principle of the present invention for commercial information processing, the fifth embodiment of the present invention provides an example of an APP interface screenshot of a personal user performing commercial information processing. Figure 5 , Figure 5 Screenshot of the APP interface for processing business needs input by individual users provided in Example 5 of the present invention; see Figure 6 , Figure 6 This is a screenshot of the APP interface for analyzing business needs, generating solutions, and matching business cards with partners, as provided in Example 5 of the present invention.

[0114] Figure 5 Zhongwei has realized the business card of individual user Zhang Xiaozhang through the above embodiment, which includes the individual user's name, company, position, user description, and core competence. In the screenshot of this interface, the user can be configured to input the user's demand information, use the intelligent matching system to analyze the demand, generate cooperation opportunities, find suitable partners, and achieve the satisfaction of the demand. Figure 5 You can input the user's business project demand data (core business / products / resources, and user needs), and after triggering the "Generate Cooperation Point" button, match the appropriate partner to meet the needs.

[0115] Figure 6 The design plan generated by analyzing the user input requirement "I need to renovate my house" through the above embodiment of the present invention is divided into three sub-steps (1. House style setting; 2. Purchase decoration materials; 3. Determine the construction team for renovation), and the business card data of potential partners is matched in each sub-step. Figure 6 In the next step, the business cards of two people are matched in the sub-step of agreeing on the house style; in the sub-step of purchasing decoration materials, the business cards of two people are matched. The user only needs to contact the matching business card to form a cooperation, which is called the network configuration solution.

[0116] The method flow implementation solutions in the above embodiments can be implemented as software programs or packaged as software modules. Persons of ordinary skill in the art can understand and implement these solutions as various software programs, mobile apps, APIs, or SaaS (Software as a Service) services without inventive effort. The method flow implementation solutions in the embodiments of the present invention, embodied in software programs or packaged as software modules, also fall within the spirit and scope of the present invention.

[0117] The module-based embodiments described in the above embodiments are merely illustrative, wherein the units described as separate components may or may not be physically separate. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiments. Those skilled in the art can understand and implement the present embodiments without inventive effort.

[0118] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for processing commercial information, comprising: Obtaining business card information input by a user, and generating first business card data of the user through an artificial intelligence model; Obtaining business project demand data input by the user; Generating business project solution data based on the business project demand data; Matching at least one second business card data using a business activity artificial intelligence model according to the business project plan data to form a second business card data set corresponding to the user's business project needs; Business information processing between the first business card data and the second business card data is established based on the second business card data set.

2. The method according to claim 1, characterized in that Further including: Matching at least one third business card data based on the user's capability information and / or the user's business resource information and / or the business product / service information provided by the user and / or the user's demand information in the first business card data to form a third business card data set corresponding to the first business card data; Based on the third business card data set, business information processing between the first business card data and the third business card data is established.

3. The method according to claim 2, characterized in that The matching of at least one third business card data includes: Obtain business card data, user business friend information, user behavior data, and user chat content data; Analyze and generate social relationship graphs based on the business social network analysis model; Generate user interest vector based on user behavior model; In the hybrid recommendation model, at least one third business card data is generated according to the social relationship graph and the user interest vector.

4. The method according to claim 1, wherein The acquiring of business card information input by the user and generating first business card data of the user by using an artificial intelligence model includes: Obtaining user identity information and / or business card style information and / or user capability information and / or user business resource information and / or business product / service information provided by the user and / or user demand information input by the user; Invoking a text artificial intelligence model to generate text for the user's first business card; Invoking an image artificial intelligence model to generate an image of the user's first business card; The first business card data of the user is constructed according to the text in the generated first business card and the picture in the first business card.

5. The method according to claim 1, wherein Generating business project solution data based on the business project demand data includes: Obtaining description information of the business project requirements; Invoke the business project implementation model to generate business solution sub-steps to achieve the business project requirements; In the business plan sub-step, second business card data and / or business cooperation plan data corresponding to the business plan sub-step are generated based on the user's capability information and / or the user's business resource information and / or the business product / service information provided by the user.

6. The method according to claim 1, characterized in that Further including: Acquire corporate business card information input by a corporate user, and generate first corporate business card data of the corporate user; The first enterprise business card data is associated with at least one first business card data; Obtain enterprise knowledge base training data from enterprise users and generate enterprise product and service knowledge base models; The product and service demand information of the second business card data is responded to according to the enterprise product and service knowledge base model.

7. The method according to claim 6, characterized in that The product service demand information in response to the second business card data based on the enterprise product service knowledge base model includes: generating a business plan sub-step of implementing the product and service requirements of the second business card data based on the product and service requirements of the second business card data as input; Using the business cases and / or industry rules in the knowledge base model and taking the business solution sub-steps as input, business cooperation solution data corresponding to the business solution sub-steps are generated.

8. The method according to claim 1, characterized in that Further including: In the business information processing between the first business card data and the second business card data, the business contract data generated by the contract macro model according to the business project demand data and the business project plan data is loaded; The business contract data is edited and signed based on the first business card data and the second business card data.

9. The method according to claim 1, characterized in that The step of matching at least one second business card data with a business activity artificial intelligence model according to the business project plan data to form a second business card data set corresponding to the business project requirements of the user includes: Performing semantic analysis on the business project plan data using a business activity artificial intelligence model to generate a structured semantic architecture; Extracting domain entities from the structured semantic architecture; A query condition is constructed based on the extracted domain entity, and a second business card data candidate set that meets the condition is retrieved from the business activity knowledge graph.

10. The method according to claim 9, characterized in that Further including: Collect data from business card data and / or corporate business information and / or product and service description information and / or user social content information and / or commercial contract transaction information; Conduct mixed extraction of business activity knowledge; Integrate the business activity knowledge of upstream and downstream business activity entities to form a global business activity knowledge map.

Images