A material recommendation method, device, equipment and medium

Abstract

The application belongs to the field of big data, and relates to a material recommendation method, comprising the following steps: obtaining conversation archive information between a financial terminal and a client terminal, extracting keywords of the conversation archive information to obtain conversation keywords of the conversation archive information; when receiving a participation request sent by the client terminal, obtaining a target material delivery strategy from a preset material delivery library; matching the conversation keywords with material keywords of material information; if the matching is successful, generating a to-be-recommended material based on the material information, determining a delivery time based on a conversation time and a time condition in the conversation archive information; and sending the to-be-recommended material and the delivery time to the financial terminal, so that the financial terminal sends the to-be-recommended material to the client terminal based on the delivery time. The application also provides a device, equipment and medium. In addition, the application also relates to blockchain technology, and the conversation archive information and the to-be-recommended material can be stored in the blockchain. The application can improve the recommendation accuracy of the to-be-recommended material.

Description

Technical Field

[0001] The present application relates to the field of big data technology, and in particular to a material recommendation method, device, equipment and medium. Background Art

[0002] The rapid development of the financial industry and the diversification and deepening of banking services have placed higher demands on customer service. Financial managers, in particular, play a crucial role in customer marketing and service optimization. In recent years, to enhance customer retention and improve marketing efficiency, many banks have adopted instant messaging as a primary channel for communicating with customers. By adding customers as friends, they establish a one-on-one personalized communication platform. Financial managers leverage this opportunity to deliver customized information about financial products and services tailored to their specific needs. This model has become a key component of modern banking marketing.

[0003] However, traditional instant messaging marketing methods have exposed significant efficiency bottlenecks in practice. Specifically, financial managers must spend considerable time reading and analyzing customer conversations to accurately grasp their potential needs. Furthermore, they must manually search through a vast database of campaign materials to find the most suitable marketing materials for each client. This process is not only extremely tedious and time-consuming, but also relies heavily on the financial manager's personal experience and professional judgment, making it difficult to effectively ensure the accuracy and efficiency of the promotional materials, making it difficult to achieve precise and efficient promotional materials. Summary of the Invention

[0004] The purpose of the embodiments of the present application is to propose a material recommendation method, device, equipment and medium to solve the existing problem of difficulty in achieving accurate and efficient material push.

[0005] In order to solve the above technical problems, the present application provides a material recommendation method, which adopts the following technical solutions:

[0006] The session archive information between the financial management terminal and the client terminal is obtained, and the keywords of the session archive information are extracted to obtain the session keywords of the session archive information; when a participation request sent by the client terminal is received, the target material delivery strategy is obtained from the preset material delivery library, and the participation request is triggered by the client terminal replying to the activity invitation information of the financial management terminal. The target material delivery strategy includes time conditions, matching rules and material information; based on the matching rules, the session keywords are matched with the material keywords of the material information; if the match is successful, the material to be recommended is generated based on the material information, and the delivery time of the material to be recommended is determined based on the session time and time conditions in the session archive information; the material to be recommended and the delivery time are sent to the financial management terminal, so that the financial management terminal sends the material to be recommended to the client terminal based on the delivery time.

[0007] Furthermore, the step of obtaining the target material delivery strategy from the preset material delivery library specifically includes:

[0008] Based on conversation keywords, a conversation scenario feature vector of conversation archive information is constructed; a preset material delivery library is obtained, which includes multiple material delivery strategies and the strategy scenario feature vector corresponding to each material delivery strategy; the similarity between the conversation scenario feature vector and the strategy scenario feature vector is calculated to obtain a similarity result; based on the similarity result, a target material delivery strategy is determined from multiple material delivery strategies.

[0009] Furthermore, the matching rule includes a fuzzy matching algorithm. Based on the matching rule, the steps of matching the conversation keywords with the material keywords of the material information specifically include:

[0010] Extract keywords of the material information to obtain material keywords of the material information; calculate the similarity score between the conversation keywords and the material keywords based on the fuzzy matching algorithm; if the similarity score reaches a preset threshold, it is determined that the conversation keywords and the material keywords are successfully matched; if the similarity score is less than the threshold, it is determined that the conversation keywords and the material keywords are not successfully matched.

[0011] Furthermore, the step of extracting keywords from the session archive information to obtain session keywords from the session archive information specifically includes:

[0012] The session archive information is preprocessed to obtain target session archive information; a preset word segmentation algorithm is used to segment the target session archive information to obtain multiple word segments; a weight value of each word segment is calculated according to a preset keyword extraction rule, and based on the weight value, a session keyword is determined from the multiple word segments.

[0013] Furthermore, the step of generating the recommended materials based on the material information specifically includes:

[0014] Using text mining algorithms, we conduct intent analysis on the target conversation archive information to obtain the conversation intent information of the customer terminal. Based on the conversation intent information, we obtain product information of matching adapted products from the established product library. Based on the product information and material information, we generate materials to be recommended.

[0015] Furthermore, based on the session time and time conditions in the session archive information, the step of determining the release time of the recommended material specifically includes:

[0016] The session time in the session archive information is parsed to obtain the session start time and session end time of the session time; based on the time condition, the session start time and the session end time, a preset algorithm is used to calculate the release time of the recommended material.

[0017] Furthermore, after the step of sending the recommended materials and the delivery time to the financial terminal so that the financial terminal sends the recommended materials to the client terminal based on the delivery time, the method further includes:

[0018] Obtain first feedback data from a client terminal regarding the recommended material; retrieve at least one historical material similar to the material to be recommended from a preset material delivery history database, and extract second feedback data of at least one historical material within a preset historical time period; and adjust the material delivery strategy in the material delivery library based on the first feedback data and the second feedback data.

[0019] In order to solve the above technical problems, the present application also provides a material recommendation device, which adopts the following technical solutions:

[0020] An acquisition module is used to acquire the session archive information between the financial management terminal and the client terminal, extract keywords from the session archive information, and obtain session keywords from the session archive information;

[0021] The receiving module is used to obtain the target material delivery strategy from the preset material delivery library when receiving a participation request sent by the client terminal. The participation request is triggered by the client terminal replying to the activity invitation information of the financial terminal. The target material delivery strategy includes time conditions, matching rules and material information;

[0022] A matching module, used to match conversation keywords with material keywords of material information based on matching rules;

[0023] A generation module is used to generate a to-be-recommended material based on the material information if the match is successful, and determine the release time of the to-be-recommended material based on the session time and time conditions in the session archive information;

[0024] The sending module is used to send the materials to be recommended and the delivery time to the financial terminal, so that the financial terminal sends the materials to be recommended to the customer terminal based on the delivery time.

[0025] In order to solve the above technical problems, an embodiment of the present application also provides a computer device, including a memory and a processor, wherein the memory stores computer-readable instructions, and when the processor executes the computer-readable instructions, the steps of the above-mentioned material recommendation method are implemented.

[0026] In order to solve the above technical problems, an embodiment of the present application also provides a computer-readable storage medium, which stores computer-readable instructions. The computer-readable instructions can be executed by at least one processor to enable at least one processor to perform the steps of the above-mentioned material recommendation method.

[0027] Compared with the prior art, the embodiments of the present application mainly have the following beneficial effects: by obtaining the conversation archive information between the financial management terminal and the customer terminal, and accurately extracting keywords, a data basis is provided for subsequent material recommendations. When a customer responds to an activity invitation, the material delivery strategy that best matches the current conversation content can be quickly matched from the preset material delivery library. Through matching rules, the precise correspondence between conversation keywords and material keywords ensures the pertinence and relevance of the materials to be recommended. At the same time, combined with the conversation time and the time conditions for material delivery, the optimal delivery time can be intelligently determined, thereby improving the efficiency and effectiveness of the delivery of materials to be recommended. Ultimately, the financial management terminal can accurately and timely send the materials to be recommended to the customer terminal based on the provided delivery time and materials to be recommended, thereby improving the accuracy of the recommendation of materials to be recommended. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the solutions in this application, a brief introduction will be given below to the drawings required for use in the description of the embodiments of this application. Obviously, the drawings described below are some embodiments of this application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0029] Figure 1 is an exemplary system architecture diagram to which the present application may be applied;

[0030] Figure 2 This is a flow chart of a material recommendation method provided in this application;

[0031] Figure 3 This is a structural diagram of a material recommendation device provided by this application;

[0032] Figure 4 This is a structural diagram of a computer device provided by this application. DETAILED DESCRIPTION

[0033] Unless otherwise defined, all technical and scientific terms used herein have the same meanings as commonly understood by those skilled in the art to which this application belongs. The terms used in the specification of the application are for the purpose of describing specific embodiments only and are not intended to limit this application. The terms "including" and "having" and any variations thereof in the specification and claims of this application and the above-mentioned drawings are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of this application or the above-mentioned drawings are used to distinguish different objects, not to describe a specific order.

[0034] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute an independent or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0035] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.

[0036] like Figure 1 As shown, system architecture 100 may include a terminal device 101, a network 102, and a server 103. Terminal device 101 may be a laptop computer 1011, a tablet computer 1012, or a mobile phone 1013. Network 102 is a medium for providing a communication link between terminal device 101 and server 103. Network 102 may include various connection types, such as wired or wireless communication links or fiber optic cables.

[0037] The user can use the terminal device 101 to interact with the server 103 via the network 102 to receive or send messages, etc. Various communication client applications can be installed on the terminal device 101, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0038] The terminal device 101 can be various electronic devices with a display screen and supporting web browsing. In addition to the laptop computer 1011, tablet computer 1012 or mobile phone 1013, the terminal device 101 can also be an e-book reader, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 (Moving Picture Experts Group Audio Layer IV) player, a laptop computer and a desktop computer, etc.

[0039] The server 103 may be a server that provides various services, such as a background server that provides support for web pages displayed on the terminal device 101 .

[0040] It should be noted that the material recommendation method provided in the embodiment of the present application is generally executed by a server, and accordingly, the material recommendation device is generally set in the server.

[0041] It should be understood that Figure 1The number of terminal devices, networks and servers in the embodiment is merely illustrative. Any number of terminal devices, networks and servers may be provided as required.

[0042] Continue to refer Figure 2 , shows a flow chart of an embodiment of a material recommendation method according to the present application. The material recommendation method comprises the following steps:

[0043] Step S201: acquiring session archive information between the financial management terminal and the client terminal, extracting keywords from the session archive information, and obtaining session keywords from the session archive information.

[0044] In this embodiment, the electronic device on which the material recommendation method is executed (eg Figure 1 The server shown in FIG. 1 may obtain the text to be processed via a wired connection or a wireless connection. It should be noted that the wireless connection may include, but is not limited to, 3G / 4G / 5G connection, WiFi connection, Bluetooth connection, WiMAX connection, Zigbee connection, UWB (ultra wideband) connection, and other wireless connection methods currently known or to be developed in the future.

[0045] Financial management terminals refer to terminal devices used to provide financial management service information, display financial management products, and manage customer financial management accounts. Financial management terminals are used for financial management-related interactions and communication with customers. These terminals include, but are not limited to, smartphones, tablets, and personal computers.

[0046] The term "client terminal" refers to the terminal device used by clients to receive information from a financial management terminal, submit financial management requests, check financial management status, and perform other operations. Client terminals include, but are not limited to, smartphones, tablet computers, and personal computers.

[0047] Conversation archive information refers to a detailed record of all interactive conversations between the financial management terminal and the client terminal within instant messaging tools. Conversation archive information may include, but is not limited to, text, voice, images, and other information. This conversation archive information can be used for subsequent analysis, processing, and as a basis for service improvement.

[0048] Keywords and conversation keywords refer to words or phrases with specific meaning or significant value extracted from conversation archives. These keywords summarize and reflect the main content and intent of conversation archives. They are used for subsequent matching, analysis, and recommendation.

[0049] Step S202: When a participation request is received from a client terminal, a target material delivery strategy is obtained from a preset material delivery library. The participation request is triggered by the client terminal replying to the activity invitation information of the financial management terminal. The target material delivery strategy includes time conditions, matching rules and material information.

[0050] Among them, the participation request refers to a response behavior actively initiated by the client terminal based on the client's trigger operation after receiving the activity invitation information sent by the financial management terminal, indicating that the client is willing to participate in the activity and triggering the execution of the subsequent recommended material delivery logic.

[0051] The event invitation information refers to information sent by the financial terminal to the client terminal to invite the client to participate in a specific event (such as financial product promotion, investment seminar, etc.). The event invitation information may include key content such as event details and participation methods.

[0052] The material delivery library refers to a database that stores various material delivery strategies. These strategies may include, but are not limited to, time conditions, matching rules, and material information for different customer groups and activity scenarios.

[0053] The target material delivery strategy is a material delivery strategy that is selected from the material delivery library based on the conditions triggered by the participation request and best matches the current event invitation information and customer profile. This target material delivery strategy can include specific time conditions, matching rules, and material information.

[0054] Among them, time conditions: a set of rules or conditions defined in the target material delivery strategy, used to determine the specific delivery time of the recommended material, such as the delivery start time, end time, delivery period, etc.

[0055] Matching rules are rules or algorithms used to compare and match conversation keywords with material keywords in material information to determine the relevance or degree of match between the two. They are an important basis for deciding whether to recommend a material to a customer. Examples include exact matching algorithms (a successful match is considered only when the conversation keyword and the material keyword are exactly the same), fuzzy matching algorithms (conversation keywords and material keywords are semantically similar or related, such as using synonyms, antonyms, or related phrases for matching), and partial matching algorithms (a successful match is considered as long as the conversation keyword contains a portion of the material keyword. For example, if the material keyword is "investment product" and the conversation keyword is "high-yield investment product," the partial matching rule will consider them a match).

[0056] The material information refers to specific content of a material to be recommended to the customer specified in the target material delivery strategy. The material information can include but is not limited to material type (such as text, picture, video, etc.), material title, material description, material keyword, etc.

[0057] In step S203, the session keyword is matched with the material keyword of the material information based on the matching rule.

[0058] The material keyword refers to a keyword or phrase used to describe the content, theme or characteristics of the material extracted or defined in the material information. The material keyword is used to match the session keyword to evaluate the relevance of the material to the customer demand.

[0059] In an example, if the matching rule is an exact matching algorithm, the session keyword and the material keyword are completely the same to be considered as a successful match. If the matching rule is a fuzzy matching algorithm, the session keyword and the material keyword are similar or related in semantics, such as using synonyms, near synonyms or related phrases for matching. If the matching rule is a partial matching algorithm, as long as the session keyword contains a part of the material keyword, it is considered as a successful match. For example, if the material keyword is "investment product" and the session keyword is "high-yield investment product", the partial matching rule considers them as a match.

[0060] In step S204, if the matching is successful, the material to be recommended is generated based on the material information, and the delivery time of the material to be recommended is determined based on the session time and the time condition in the session archive information.

[0061] The material to be recommended refers to the specific material content selected from the material information and prepared to be recommended to the customer based on the successful matching of the session keyword and the material keyword according to the matching rule.

[0062] The session time refers to the occurrence time of a specific interactive session between the financial terminal and the customer terminal recorded in the session archive information. The session time can be used to determine the delivery time of the material to be recommended in combination with the time condition.

[0063] The delivery time refers to the specific time point or time period at which the material to be recommended is actually sent to the customer terminal determined according to the time condition and the session time. The delivery time is one of the key factors to ensure the delivery effect of the material to be recommended.

[0064] In step S205, the material to be recommended and the delivery time are sent to the financial terminal, so that the financial terminal sends the material to be recommended to the customer terminal based on the delivery time.

[0065] In one embodiment, the data of the materials to be recommended and the delivery time can be packaged to generate a data packet, which is then sent to the financial terminal through a secure communication protocol. The financial terminal can have a built-in time-sensitive distribution engine. The engine can parse the received data packet, automatically schedule the materials to be recommended according to the delivery time, and accurately push the materials to be recommended to the customer terminal to achieve personalized and timely material display. Taking the promotion of a certain financial management product as an example, assuming that the material to be recommended is "Introduction to High-yield Financial Management Products", the delivery time is set to 9 to 10 am on the working day closest to the end of the session time. Through the embodiment of the present application, the material to be recommended "Introduction to High-yield Financial Management Products" and the delivery time are sent to the financial terminal. After receiving the instruction, the financial terminal pushes the material to be recommended "Introduction to High-yield Financial Management Products" to the customer terminal on time from 9 to 10 am on the working day closest to the end of the session.

[0066] The embodiment of the present application can obtain the conversation archive information between the financial management terminal and the client terminal, and accurately extract keywords, to provide a data basis for subsequent material recommendations. When a customer responds to an activity invitation, the material delivery strategy that best matches the current conversation content can be quickly matched from the preset material delivery library. Through matching rules, the precise correspondence between conversation keywords and material keywords ensures the pertinence and relevance of the materials to be recommended. At the same time, combined with the conversation time and the time conditions for material delivery, the optimal delivery time can be intelligently determined, thereby improving the efficiency and effectiveness of the delivery of materials to be recommended. Ultimately, the financial management terminal can accurately and timely send the materials to be recommended to the client terminal based on the provided delivery time and the materials to be recommended, thereby improving the accuracy of the recommendation of the materials to be recommended.

[0067] In some optional implementations of this embodiment, step 201, extracting keywords from the session archive information to obtain session keywords from the session archive information, specifically includes the following steps:

[0068] The session archive information is preprocessed to obtain target session archive information; a preset word segmentation algorithm is used to segment the target session archive information to obtain multiple word segments; a weight value of each word segment is calculated according to a preset keyword extraction rule, and based on the weight value, a session keyword is determined from the multiple word segments.

[0069] In one example, we'll use conversation archive information from the financial sector as an example. First, preprocess the financial conversation archive information. This conversation archive information may come from phone recordings between customers and customer service personnel, online chat logs, or video conference recordings, and contains a large amount of unstructured data. Preprocessing steps include noise removal (such as silence and background music), speech recognition (converting audio to text), and data cleaning (removing irrelevant characters and standardizing the format), thereby obtaining the target conversation archive information. Next, a preset word segmentation algorithm is used to segment the target conversation archive information. In the financial sector, the word segmentation algorithm must consider the integrity of financial terminology, such as "stock trading" and "interest rate adjustment," to avoid incorrect segmentation of specialized terms. This embodiment can use a word segmentation model based on conditional random fields (CRF). This model, trained on a large financial corpus, can effectively identify and retain specialized financial terms. Subsequently, a weight value is calculated for each word segmentation based on preset keyword extraction rules. The rule design combines Term Frequency-Inverse Document Frequency (TF-IDF) with the knowledge graph in the financial field. It not only considers the frequency of word occurrence in the conversation, but also evaluates its importance in the financial knowledge system. For example, the word "stock" will have a higher weight than common words such as "hello" and "thank you" in conversations that frequently discuss market dynamics. Finally, based on the weight value, the conversation keywords are determined from multiple word segments. Accurate extraction is achieved by setting a threshold or selecting the top N word segments as keywords. For example, in a conversation about mortgage interest rate adjustments, keywords may include "mortgage", "interest rate", "adjustment", etc. These keywords directly reflect the core content of the conversation archive information.

[0070] This embodiment of the present application preprocesses conversation archive information to remove noise and irrelevant characters, ensuring the accuracy of subsequent word segmentation and keyword extraction. The pre-set word segmentation algorithm accurately identifies financial terminology, avoiding the loss or misunderstanding of key information due to word segmentation errors. By accurately extracting conversation keywords, a more precise understanding of customer needs and concerns can be achieved.

[0071] In some optional implementations of this embodiment, in step S202, a target material delivery strategy is obtained from a preset material delivery library, specifically including the following steps:

[0072] Based on conversation keywords, a conversation scenario feature vector of conversation archive information is constructed; a preset material delivery library is obtained, which includes multiple material delivery strategies and the strategy scenario feature vector corresponding to each material delivery strategy; the similarity between the conversation scenario feature vector and the strategy scenario feature vector is calculated to obtain a similarity result; based on the similarity result, a target material delivery strategy is determined from multiple material delivery strategies.

[0073] In one example, the conversation archive information in the financial field can be used as an example for illustration. After preprocessing the conversation archive information, including noise removal, word segmentation, and keyword extraction, each extracted conversation keyword is encoded into a unique digital identifier according to a preset vocabulary. Then, based on the occurrence of each conversation keyword, a conversation scenario feature vector is constructed. The preset material delivery library contains multiple material delivery strategies and corresponding strategy scenario feature vectors. The similarity between the conversation scenario feature vector and the strategy scenario feature vector is calculated to obtain a similarity result. Common methods such as cosine similarity and Euclidean distance can be used for similarity calculation. Based on the similarity result, the strategy that best matches the target conversation scenario is selected from multiple material delivery strategies as the target material delivery strategy.

[0074] The embodiment of the present application can accurately capture the needs and preferences expressed by customers in the conversation by extracting conversation keywords and constructing conversation scenario feature vectors. This keyword-based feature extraction method can deeply explore customer intentions and provide strong support for the selection of subsequent material delivery strategies. At the same time, similarity calculation with the strategy scenario feature vectors in the preset material delivery library can further ensure the accurate match between the material delivery strategy and customer needs. Traditional material delivery methods often rely on manual judgment and experience-based decision-making, which is inefficient and difficult to guarantee accuracy. The solution provided by this embodiment can improve the efficiency of material delivery through automated feature extraction, similarity calculation and strategy selection processes. The processing and analysis of a large amount of conversation archive information can be completed in a short period of time, and the target material delivery strategy can be quickly determined, thereby achieving timely response and effective satisfaction of customer needs.

[0075] In some optional implementations of this embodiment, the matching rule includes a fuzzy matching algorithm. Step S203, based on the matching rule, matches the conversation keyword with the material keyword of the material information, specifically including the following steps:

[0076] Extract keywords of the material information to obtain material keywords of the material information; calculate the similarity score between the conversation keywords and the material keywords based on the fuzzy matching algorithm; if the similarity score reaches a preset threshold, it is determined that the conversation keywords and the material keywords are successfully matched; if the similarity score is less than the threshold, it is determined that the conversation keywords and the material keywords are not successfully matched.

[0077] In one example, key information, such as the financial product name, type, interest rate, and term, can be extracted from the material information as material keywords. A fuzzy matching algorithm is used to calculate the similarity score between the conversation keywords in the session archive and the material keywords. This fuzzy matching algorithm can consider factors such as semantic similarity and spelling errors to improve matching accuracy. A preset threshold can be set. If the similarity score reaches or exceeds this threshold, the conversation keywords are considered a match with the material keywords, triggering the corresponding material release strategy. If the similarity score falls below the threshold, the match is considered unsuccessful, and the corresponding material release strategy is not triggered.

[0078] The embodiment of the present application can more accurately identify the content related to material information in the user session by extracting the keywords of the material information and performing fuzzy matching with the conversation keywords. The application of the fuzzy matching algorithm makes it possible to match the conversation keywords based on the similarity between the conversation keywords and the material keywords even if the conversation keywords of the session archive information are not completely consistent with the material keywords in the material information, thereby improving the accuracy of the match. The application of the fuzzy matching algorithm enables the solution provided by this embodiment to cope with a variety of user conversations and material information expressions. Whether it is a concise word or a complex sentence entered by the customer, effective keyword extraction and matching can be performed.

[0079] In some optional implementations of this embodiment, step S204, generating materials to be recommended based on the material information, specifically includes the following steps:

[0080] Using text mining algorithms, we conduct intent analysis on the target conversation archive information to obtain the conversation intent information of the customer terminal. Based on the conversation intent information, we obtain product information of matching adapted products from the established product library. Based on the product information and material information, we generate materials to be recommended.

[0081] In one example, text mining algorithms (such as TF-IDF, Word2Vec, or BERT) can be used to analyze the intent of target conversation archives to identify customer conversational intent, such as inquiries, consultations, and investment intentions regarding financial products. Based on this conversational intent, product information for compatible products matching the customer's needs is retrieved from a pre-existing library of financial products (including but not limited to wealth management products, loan products, and insurance products). Recommended items are generated by combining the product information with the compatible product information. For example, the BERT text mining algorithm can be used to analyze the intent of target conversation archives from a client terminal to identify the customer's inquiry intent for short-term, low-risk wealth management products. Based on the identified conversational intent, wealth management products that meet the "short-term, low-risk" criteria are screened from the financial product library, such as a money market fund or a short-term fixed-term wealth management product. Recommended items are generated by combining the product information and the product information of these wealth management products that meet the "short-term, low-risk" criteria.

[0082] The embodiments of the present application can accurately analyze the conversation intention information of the customer terminal to quickly understand the customer's needs and provide more personalized material recommendations. Customers do not need to screen through a large number of products themselves, saving time and energy.

[0083] In some optional implementations of this embodiment, in step S204, based on the session time and time conditions in the session archive information, the release time of the recommended material is determined, which specifically includes the following steps:

[0084] The session time in the session archive information is parsed to obtain the session start time and session end time of the session time; based on the time condition, the session start time and the session end time, a preset algorithm is used to calculate the release time of the recommended material.

[0085] In one example, the start and end times of the session corresponding to the session archive information can be extracted. The time condition is set to 9:00 AM to 10:00 AM on the weekday closest to the session end time. Then, a preset algorithm (e.g., a time window algorithm) is used to calculate a reasonable time range based on the session end time, so that this time range falls within the range of 9:00 AM to 10:00 AM on the weekday closest to the session end time.

[0086] This embodiment of the application can analyze conversation time to understand when customers consult or inquire, thereby more accurately grasping their interests and needs. Combined with a preset algorithm, recommended products can be delivered during the time period when customers are most likely to be interested in or need product information, improving the accuracy and effectiveness of recommendations.

[0087] In some optional implementations of this embodiment, after the recommended materials and the delivery time are sent to the financial management terminal in step S205 so that the financial management terminal sends the recommended materials to the client terminal based on the delivery time, the following steps are further specifically included:

[0088] Obtain first feedback data from a client terminal regarding the recommended material; retrieve at least one historical material similar to the material to be recommended from a preset material delivery history database, and extract second feedback data of at least one historical material within a preset historical time period; and adjust the material delivery strategy in the material delivery library based on the first feedback data and the second feedback data.

[0089] In one example, after a client terminal receives a material to be recommended, first feedback data from the client terminal can be obtained. Historical materials similar to the material to be recommended to the client terminal (such as the time deposit product details page and fund investment guide previously recommended to other client terminals) are retrieved from the material delivery history database, and second feedback data (such as click-through rate, conversion rate, etc.) of these historical materials within a preset historical time period are extracted. Based on the first feedback data of the client terminal and the second feedback data of similar historical materials, the material delivery strategy in the material delivery library is adjusted, such as adjusting time conditions, optimizing matching rules, improving material information, etc., to improve the accuracy and effectiveness of subsequent material recommendations.

[0090] The first feedback data in the embodiments of this application directly reflects the client's immediate response to the currently recommended item. This data provides a direct basis for evaluating the effectiveness of the recommendation. Based on a comprehensive analysis of the first and second feedback data, the material placement strategy can be adjusted in a timely manner, such as adjusting time conditions, optimizing matching rules, and improving the content of material information, thereby improving the accuracy and effectiveness of the recommendation.

[0091] It should be emphasized that in order to further ensure the privacy and security of the above-mentioned session archive information, session keywords, material placement strategies in the material placement library, and materials to be recommended, the above-mentioned session archive information, session keywords, material placement strategies in the material placement library, and materials to be recommended can also be stored in a blockchain node.

[0092] The blockchain referred to in this application is a new application model for computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanisms, and encryption algorithms. Blockchain is essentially a decentralized database, a series of data blocks generated using cryptographic methods. Each data block contains information about a batch of network transactions, which is used to verify the validity of this information (to prevent counterfeiting) and generate the next block. Blockchain can include the underlying blockchain platform, the platform product service layer, and the application service layer.

[0093] The embodiments of the present application can obtain and process the conversation archive information, the conversation keywords, the target material delivery strategy, the material keywords, the conversation time, the material information, and the to-be-recommended material based on the related algorithm of the natural language processing technology (NLP) under artificial intelligence. The artificial intelligence (AI) is to use a digital computer or a machine controlled by a digital computer to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use the knowledge to obtain the best results.

[0094] The natural language processing technology is a key branch in the field of artificial intelligence, which focuses on giving computers the ability to understand and process human language, enabling them to analyze, understand, and generate natural language text like humans. By simulating some core mechanisms of human language processing, NLP technology can deeply process, analyze, and interpret text data, extract key information, understand semantic content, and execute complex tasks or generate natural language responses accordingly. The natural language processing technology can be used to extract keywords from the conversation archive information, generate conversation keywords of the conversation archive information, extract keywords from the material information, generate material keywords, generate to-be-recommended materials based on the material information, and generate delivery time based on the conversation time and time conditions.

[0095] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by computer readable instructions instructing related hardware, and the computer readable instructions can be stored in a computer readable storage medium. When the program is executed, it can include the processes of the above-mentioned embodiments of each method. The storage medium can be a non-volatile storage medium such as a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0096] It should be understood that although each step in the flowchart of the accompanying drawings is displayed in sequence according to the direction of the arrow, these steps are not necessarily executed in sequence according to the direction of the arrow. Unless otherwise specified herein, the execution of these steps is not strictly limited in sequence, and they can be executed in other orders. Moreover, at least part of the steps in the flowchart of the accompanying drawings can include multiple sub-steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution sequence is not necessarily sequential, but can be executed alternately or alternately with at least part of other steps or sub-steps or stages of other steps.

[0097] Further reference Figure 3 , as a response to the above Figure 2 The present application provides an embodiment of a material recommendation device. Figure 2 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0098] like Figure 3 As shown, the material recommendation device 400 of this embodiment includes: an acquisition module 401, a receiving module 402, a matching module 403, a generation module 404 and a sending module 405. Among them:

[0099] The acquisition module 401 is used to acquire the session archive information between the financial management terminal and the client terminal, extract the keywords of the session archive information, and obtain the session keywords of the session archive information;

[0100] Receiving module 402 is configured to obtain a target material delivery strategy from a preset material delivery library upon receiving a participation request from a client terminal. The participation request is triggered by the client terminal replying to the activity invitation information from the financial terminal. The target material delivery strategy includes time conditions, matching rules, and material information.

[0101] Matching module 403, for matching conversation keywords with material keywords of material information based on matching rules;

[0102] A generation module 404 is configured to generate a material to be recommended based on the material information if the match is successful, and determine a release time for the material to be recommended based on the session time and time conditions in the session archive information;

[0103] The sending module 405 is used to send the materials to be recommended and the delivery time to the financial terminal, so that the financial terminal sends the materials to be recommended to the client terminal based on the delivery time.

[0104] The embodiment of the present application can obtain the conversation archive information between the financial management terminal and the client terminal, and accurately extract keywords, to provide a data basis for subsequent material recommendations. When a customer responds to an activity invitation, the material delivery strategy that best matches the current conversation content can be quickly matched from the preset material delivery library. Through matching rules, the precise correspondence between conversation keywords and material keywords ensures the pertinence and relevance of the materials to be recommended. At the same time, combined with the conversation time and the time conditions for material delivery, the optimal delivery time can be intelligently determined, thereby improving the efficiency and effectiveness of the delivery of materials to be recommended. Ultimately, the financial management terminal can accurately and timely send the materials to be recommended to the client terminal based on the provided delivery time and the materials to be recommended, thereby improving the accuracy of the recommendation of the materials to be recommended.

[0105] In one embodiment, the acquisition module 401 includes:

[0106] a preprocessing submodule configured to preprocess the session archive information to obtain target session archive information;

[0107] a word segmentation submodule configured to perform word segmentation on the target session archive information by using a preset word segmentation algorithm to obtain a plurality of segmented words;

[0108] a weight calculation submodule configured to calculate a weight value of each segmented word according to a preset keyword extraction rule, and determine a session keyword from the plurality of segmented words based on the weight value.

[0109] The embodiments of the present application can preprocess the session archive information to remove noise and irrelevant characters, thereby ensuring the accuracy of subsequent word segmentation and keyword extraction. The preset word segmentation algorithm can accurately identify financial professional vocabulary, thereby avoiding loss or misunderstanding of key information due to word segmentation errors. By accurately extracting session keywords, customer needs and concerns can be more accurately understood.

[0110] In an embodiment, the receiving module 402 includes:

[0111] a construction submodule configured to construct a session scenario feature vector of the session archive information based on the session keyword;

[0112] a first obtaining submodule configured to obtain a preset material launching library, the material launching library including a plurality of material launching strategies and a strategy scenario feature vector corresponding to each material launching strategy;

[0113] a similarity calculation submodule configured to calculate a similarity between the session scenario feature vector and the strategy scenario feature vector to obtain a similarity result;

[0114] a determination submodule configured to determine a target material launching strategy from the plurality of material launching strategies according to the similarity result.

[0115] The embodiments of the present application can accurately capture customer needs and preferences expressed in the session by extracting session keywords and constructing a session scenario feature vector. This keyword-based feature extraction method can deeply mine customer intentions and provide strong support for the selection of subsequent material launching strategies. At the same time, similarity calculation with the strategy scenario feature vector in the preset material launching library can further ensure the accurate matching of the material launching strategy and the customer needs. Traditional material launching methods often rely on manual judgment and experience-based decision making, which is low in efficiency and difficult to ensure accuracy. The scheme provided by the embodiments can improve the efficiency of material launching through the automatic feature extraction, similarity calculation and strategy selection process. A large amount of session archive information can be processed and analyzed in a short time, and the target material launching strategy can be quickly determined, thereby realizing timely response and effective satisfaction of customer needs.

[0116] In one embodiment, the matching module 403 includes:

[0117] The extraction submodule is used to extract the keywords of the material information and obtain the material keywords of the material information;

[0118] The score calculation submodule is used to calculate the similarity scores between conversation keywords and material keywords based on the fuzzy matching algorithm;

[0119] The first determination submodule is configured to determine whether the conversation keyword and the material keyword are successfully matched if the similarity score reaches a preset threshold;

[0120] The second determination submodule is configured to determine that the conversation keyword and the material keyword are not successfully matched if the similarity score is less than a threshold.

[0121] The embodiment of the present application can more accurately identify the content related to material information in the user session by extracting the keywords of the material information and performing fuzzy matching with the conversation keywords. The application of the fuzzy matching algorithm makes it possible to match the conversation keywords based on the similarity between the conversation keywords and the material keywords even if the conversation keywords of the session archive information are not completely consistent with the material keywords in the material information, thereby improving the accuracy of the match. The application of the fuzzy matching algorithm enables the solution provided by this embodiment to cope with a variety of user conversations and material information expressions. Whether it is a concise word or a complex sentence entered by the customer, effective keyword extraction and matching can be performed.

[0122] In one embodiment, the generating module 404 includes:

[0123] The analysis submodule is used to use text mining algorithms to perform intent analysis on the target session archive information to obtain the session intent information of the client terminal;

[0124] The second acquisition submodule is used to obtain product information of matching adapted products from the constructed product library according to the session intention information;

[0125] The generation submodule is used to generate materials to be recommended based on product information and material information.

[0126] The embodiments of the present application can accurately analyze the conversation intention information of the customer terminal to quickly understand the customer's needs and provide more personalized material recommendations. Customers do not need to screen through a large number of products themselves, saving time and energy.

[0127] In one embodiment, the generating module 404 includes:

[0128] The parsing submodule is used to parse the session time in the session archive information to obtain the session start time and session end time of the session time;

[0129] The time calculation submodule is used to calculate the delivery time of the recommended material based on the time condition, the session start time and the session end time using a preset algorithm.

[0130] This embodiment of the application can analyze conversation time to understand when customers consult or inquire, thereby more accurately grasping their interests and needs. Combined with a preset algorithm, recommended products can be delivered during the time period when customers are most likely to be interested in or need product information, improving the accuracy and effectiveness of recommendations.

[0131] In one embodiment, the material recommendation device 400 further includes:

[0132] A data acquisition module, used to acquire first feedback data of the client terminal regarding the recommended material;

[0133] A retrieval module is used to retrieve at least one historical material similar to the material to be recommended from a preset material delivery history database, and extract second feedback data of the at least one historical material within a preset historical time period;

[0134] The adjustment module is used to adjust the material delivery strategy in the material delivery library based on the first feedback data and the second feedback data.

[0135] The first feedback data in the embodiments of this application directly reflects the client's immediate response to the currently recommended item. This data provides a direct basis for evaluating the effectiveness of the recommendation. Based on a comprehensive analysis of the first and second feedback data, the material placement strategy can be adjusted in a timely manner, such as adjusting time conditions, optimizing matching rules, and improving the content of material information, thereby improving the accuracy and effectiveness of the recommendation.

[0136] To solve the above technical problems, the present application also provides a computer device. Figure 4 , Figure 4 This is a basic structural block diagram of the computer device in this embodiment.

[0137] The computer device 6 includes a memory 61, a processor 62, and a network interface 63 that are interconnected through a system bus. It should be noted that the figure only shows a computer device 6 with a memory 61, a processor 62, and a network interface 63, but it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented instead. Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to a microprocessor, an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA), a digital signal processor (DSP), an embedded device, etc.

[0138] Computer devices can be desktop computers, laptops, PDAs, cloud servers, etc. Computer devices can interact with users through keyboards, mice, remote controls, touchpads, or voice-activated devices.

[0139] Memory 61 includes at least one type of readable storage medium, including flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic storage, magnetic disk, optical disk, etc. In some embodiments, memory 61 may be an internal storage unit of computer device 6, such as the hard disk or memory of computer device 6. In other embodiments, memory 61 may also be an external storage device of computer device 6, such as a plug-in hard disk, smart memory card (SMC), secure digital (SD) card, flash memory card, etc. equipped on computer device 6. Of course, memory 61 may also include both internal storage units and external storage devices of computer device 6. In this embodiment, memory 61 is generally used to store the operating system and various application software installed on computer device 6, such as computer-readable instructions for the material recommendation method. In addition, memory 61 may also be used to temporarily store various types of data that have been output or are about to be output.

[0140] The processor 62 may, in some embodiments, be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip. The processor 62 is generally used to control the overall operation of the computer device 6. In the present embodiment, the processor 62 is used to run computer readable instructions or process data stored in the memory 61, such as computer readable instructions of the material recommendation method.

[0141] The network interface 63 can include a wireless network interface or a wired network interface, and is generally used to establish a communication connection between the computer device 6 and other electronic devices.

[0142] The embodiments of the present application can obtain the session archive information of the financial terminal and the client terminal, and accurately extract keywords, thereby providing a data basis for subsequent material recommendation. When the client responds to the invitation of the activity, the most suitable material placement strategy that conforms to the current session content can be quickly matched from the preset material placement library. Through the matching rules, the accurate correspondence between the session keywords and the material keywords ensures the pertinence and relevance of the recommended material. Meanwhile, in combination with the session time and the time condition of the material placement, the best placement time can be intelligently determined, thereby improving the efficiency and effect of the recommended material placement. Finally, the financial terminal can accurately and timely send the recommended material to the client terminal according to the provided placement time and the recommended material, thereby improving the recommendation accuracy of the recommended material.

[0143] The present application also provides another implementation, that is, a computer readable storage medium storing computer readable instructions, the computer readable instructions being executable by at least one processor to cause the at least one processor to perform the steps of the material recommendation method as described above.

[0144] The embodiments of the present application can obtain the session archive information of the financial terminal and the client terminal, and accurately extract keywords, thereby providing a data basis for subsequent material recommendation. When the client responds to the invitation of the activity, the most suitable material placement strategy that conforms to the current session content can be quickly matched from the preset material placement library. Through the matching rules, the accurate correspondence between the session keywords and the material keywords ensures the pertinence and relevance of the recommended material. Meanwhile, in combination with the session time and the time condition of the material placement, the best placement time can be intelligently determined, thereby improving the efficiency and effect of the recommended material placement. Finally, the financial terminal can accurately and timely send the recommended material to the client terminal according to the provided placement time and the recommended material, thereby improving the recommendation accuracy of the recommended material.

[0145] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods of each embodiment of the present application.

[0146] The non-Company software tools or components appearing in the embodiments of this application are merely examples and do not represent actual use.

[0147] Obviously, the embodiments described above are only some of the embodiments of the present application, rather than all of the embodiments. The preferred embodiments of the present application are given in the accompanying drawings, but they do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present application more thorough and comprehensive. Although the present application has been described in detail with reference to the aforementioned embodiments, for those skilled in the art, it is still possible to modify the technical solutions described in the aforementioned specific embodiments, or to make equivalent replacements for some of the technical features therein. Any equivalent structure made using the contents of the present application specification and the accompanying drawings, directly or indirectly used in other related technical fields, is also within the scope of patent protection of the present application.

Claims

1. A material recommendation method, characterized in that: The steps include: Acquire session archive information between the financial management terminal and the client terminal, extract keywords from the session archive information, and obtain session keywords from the session archive information; When receiving a participation request sent by the client terminal, obtaining a target material delivery strategy from a preset material delivery library, wherein the participation request is triggered by the client terminal replying to the activity invitation information of the financial terminal, and the target material delivery strategy includes time conditions, matching rules and material information; Based on the matching rule, matching the conversation keyword with the material keyword of the material information; If the match is successful, a material to be recommended is generated based on the material information, and a release time of the material to be recommended is determined based on the session time in the session archive information and the time condition; The material to be recommended and the delivery time are sent to the financial management terminal, so that the financial management terminal sends the material to be recommended to the customer terminal based on the delivery time.

2. The method according to claim 1, characterized in that The step of obtaining the target material delivery strategy from the preset material delivery library specifically includes: constructing a conversation scenario feature vector of the conversation archive information based on the conversation keywords; Obtaining a preset material delivery library, wherein the material delivery library includes multiple material delivery strategies and a strategy scenario feature vector corresponding to each material delivery strategy; Calculating the similarity between the conversation scenario feature vector and the strategy scenario feature vector to obtain a similarity result; According to the similarity result, a target material delivery strategy is determined from the multiple material delivery strategies.

3. The method according to claim 1, characterized in that The matching rule includes a fuzzy matching algorithm. The step of matching the conversation keyword with the material keyword of the material information based on the matching rule specifically includes: Extracting keywords from the material information to obtain material keywords from the material information; Calculating the similarity score between the conversation keyword and the material keyword according to the fuzzy matching algorithm; If the similarity score reaches a preset threshold, it is determined that the conversation keyword and the material keyword are successfully matched; If the similarity score is less than the threshold, it is determined that the conversation keyword and the material keyword are not successfully matched.

4. The method according to claim 1, wherein The step of extracting keywords from the conversation archive information to obtain conversation keywords from the conversation archive information specifically includes: Preprocessing the session archive information to obtain target session archive information; Using a preset word segmentation algorithm, segmenting the target conversation archive information to obtain multiple word segments; According to a preset keyword extraction rule, a weight value of each segmented word is calculated, and based on the weight value, a conversation keyword is determined from the multiple segmented words.

5. The method according to claim 4, characterized in that The step of generating the material to be recommended based on the material information specifically includes: Using a text mining algorithm to perform intent analysis on the target session archive information to obtain session intent information of the client terminal; According to the session intention information, obtain product information of matching adapted products from the constructed product library; Based on the product information and the material information, a material to be recommended is generated.

6. The method according to claim 1, characterized in that The step of determining the release time of the to-be-recommended material based on the session time in the session archive information and the time condition specifically includes: Parsing the session time in the session archive information to obtain the session start time and session end time of the session time; Based on the time condition, the session start time and the session end time, a preset algorithm is used to calculate the delivery time of the to-be-recommended material.

7. The method according to claim 1, characterized in that After the step of sending the to-be-recommended material and the delivery time to the financial management terminal so that the financial management terminal sends the to-be-recommended material to the client terminal based on the delivery time, the method further includes: Acquiring first feedback data from the client terminal regarding the material to be recommended; Retrieving at least one historical material similar to the material to be recommended from a preset material delivery history database, and extracting second feedback data of the at least one historical material within a preset historical time period; Based on the first feedback data and the second feedback data, the material delivery strategy in the material delivery library is adjusted.

8. A material recommendation device, characterized in that: include: An acquisition module, configured to acquire archived conversation information between the financial management terminal and the client terminal, extract keywords from the archived conversation information, and obtain conversation keywords from the archived conversation information; a receiving module configured to obtain a target material delivery strategy from a preset material delivery library upon receiving a participation request sent by the client terminal, wherein the participation request is triggered by the client terminal replying to the activity invitation information of the financial management terminal, and the target material delivery strategy includes a time condition, a matching rule, and material information; a matching module, configured to match the conversation keywords with the material keywords of the material information based on the matching rules; a generating module configured to generate a material to be recommended based on the material information if the match is successful, and determine a release time of the material to be recommended based on the session time in the session archive information and the time condition; The sending module is used to send the material to be recommended and the delivery time to the financial terminal, so that the financial terminal sends the material to be recommended to the client terminal based on the delivery time.

9. A computer device, characterized in that: The method comprises a memory and a processor, wherein the memory stores computer-readable instructions, and when the processor executes the computer-readable instructions, the steps of the material recommendation method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the material recommendation method according to any one of claims 1 to 7.

Images