Package recommendation method and device, electronic equipment, storage medium and program product

Abstract

The invention provides a package recommendation method and device, electronic equipment, a storage medium and a program product, and relates to the technical field of information. The method comprises the steps of obtaining a user relation graph and a user identifier of a to-be-recommended user, determining a transaction weight of the to-be-recommended user and a current package corresponding to the transaction weight according to the user relation graph and the user identifier, and recommending the to-be-recommended user according to the transaction weight and the current package based on a package recommendation model. Obtaining at least one to-be-recommended package and a recommendation weight corresponding to the to-be-recommended package; and determining at least one recommended package according to the to-be-recommended package and the recommendation weight, and recommending the recommended package to the to-be-recommended user. According to the method and the device, the current packages of the to-be-recommended users and the related users are analyzed, the recommended package is determined based on the package recommendation model according to the transaction weight and the current package, more personalized and reliable telecommunication service package recommendation is provided for the to-be-recommended users, and the package recommendation accuracy and the user satisfaction are improved.

Description

Technical Field

[0001] This disclosure relates to the field of information technology, and in particular to a method, apparatus, electronic device, computer-readable storage medium, and computer program product for recommending packages. Background Technology

[0002] With the continuous development of communication technology, telecom operators offer an increasingly diverse range of service packages to meet users' needs and preferences in various scenarios. However, most existing package recommendation systems rely on traditional recommendation algorithms, typically based solely on users' personal consumption records, historical data, or simple user segmentation. These systems primarily focus on individual user behavior patterns, failing to adequately consider the influence of users' social circles, such as interactions with friends and family and similar needs and preferences, leading to a decrease in the accuracy of recommendation results.

[0003] Therefore, improving the accuracy of package recommendation results and user satisfaction has become a key direction for current technological improvements.

[0004] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] This disclosure provides a method, apparatus, electronic device, storage medium, and program product for recommending service packages, which improves the accuracy of package recommendations and user satisfaction to at least a certain extent.

[0006] Other features and advantages of this disclosure will become apparent from the following detailed description, or may be learned in part by practice of this disclosure.

[0007] According to one aspect of this disclosure, a method for recommending service packages is provided, comprising: obtaining a user relationship graph and user identifiers of users to be recommended; determining, based on the user relationship graph and the user identifiers, a transaction weight of the user to be recommended and a current service package corresponding to the transaction weight, wherein the transaction weight includes a first weight and a second weight, the first weight indicating the importance of the current service package of the user to be recommended to the user to be recommended, and the second weight indicating the importance of the current service package of related users to the user to be recommended, wherein the related users are users who have a social relationship with the user to be recommended; based on a service package recommendation model, obtaining at least one service package to be recommended and a recommendation weight corresponding to the service package to be recommended according to the transaction weights and the current service package; determining at least one recommended service package according to the service package to be recommended and the recommendation weights, and recommending the recommended service package to the user to be recommended.

[0008] In some embodiments, prior to obtaining the user relationship graph and user identifiers, the method further includes: obtaining user identifiers based on account addresses in a blockchain transaction network; establishing social relationships between the user identifiers using knowledge graph technology; integrating user transaction data in the blockchain transaction network based on oracle and / or cross-chain bridging technology; and constructing the user relationship graph based on the social relationships and the user transaction data.

[0009] In some embodiments, determining the transaction weight of the user to be recommended and the current package corresponding to the transaction weight based on the user relationship graph and the user identifier includes: determining weight adjustment data based on the user relationship graph and the user identifier; obtaining an initial relationship weight; adjusting the initial relationship weight based on the weight adjustment data to obtain a relationship weight; and obtaining the transaction weight of the user to be recommended and the current package corresponding to the transaction weight based on the user relationship graph and the relationship weight.

[0010] In some embodiments, obtaining the transaction weight of the user to be recommended and the current package corresponding to the transaction weight based on the user relationship graph and the relationship weight includes: determining the association path related to the user to be recommended based on the user relationship graph and according to a preset social relationship hierarchy. The association path includes one or more of a first association path, a second association path, and a third association path. The first association path includes the user to be recommended and the user's current package. The second association path includes the user to be recommended, a related user, the relationship weight between the user to be recommended and the related user, and the related user's current package. The third association path includes the user to be recommended, at least two related users, the first related user among the at least two related users, and the user to be recommended. The relationship weights between recommended users, the relationship weights between two adjacent related users, and the current package of the last related user among the at least two related users are considered. The transaction weight corresponding to the current package included in the first association path is set to a first value, which is less than or equal to 1. The transaction weight corresponding to the current package included in the second association path is set to a second value, which is equal to the relationship weight between the user to be recommended and the related users in the second association path, and the relationship weight is less than the first value. The transaction weight corresponding to the current package included in the third association path is set to a third value, which is the product of all relationship weights in the third association path. The maximum value among the transaction weights corresponding to the same current package in the association path is taken as the final transaction weight.

[0011] In some embodiments, based on the package recommendation model, at least one package to be recommended and its corresponding recommendation weight are obtained according to the transaction weight and the current package. This includes: obtaining user characteristics, which include one or more of the following: user age, gender, network duration, city, total historical spending, transaction frequency, average spending, and package preference; inputting the user characteristics, the transaction weight, and the current package into the package recommendation model, and outputting at least one package to be recommended and its corresponding recommendation weight.

[0012] In some embodiments, based on the package recommendation model, at least one package to be recommended and its corresponding recommendation weight are obtained according to the transaction weight and the current package. This includes: obtaining package features, which include one or more of the following: package name, package price, package validity period, data limit, additional services, and package rating; inputting the package features, the transaction weight, and the current package into the package recommendation model, and outputting at least one package to be recommended and its corresponding recommendation weight.

[0013] In some embodiments, based on the package recommendation model, at least one package to be recommended and its corresponding recommendation weight are obtained according to the transaction weight and the current package. This includes: acquiring feature data, which includes user features and package features. The user features include one or more of the following: user age, gender, network duration, city, total historical spending, transaction frequency, average spending, and package preference. The package features include one or more of the following: package name, package price, package validity period, data limit, additional services, and package rating. The feature data, the transaction weight, and the current package are input into the package recommendation model, and at least one package to be recommended and its corresponding recommendation weight are output.

[0014] In some embodiments, the package recommendation model includes a machine learning model and / or a deep learning model.

[0015] In some embodiments, determining at least one recommended package based on the package to be recommended and the recommendation weight, and recommending the recommended package to the user to be recommended, includes: combining the packages to be recommended according to a preset strategy and the recommendation weight to obtain at least one recommended package.

[0016] In some embodiments, after determining at least one recommended package based on the package to be recommended and the recommendation weight, and recommending the recommended package to the user to be recommended, the method further includes: generating a smart contract in response to the user to be recommended's selection of the recommended package, so that the user to be recommended activates the package.

[0017] According to another aspect of this disclosure, a package recommendation device is also provided, comprising: an acquisition module for acquiring a user relationship graph and a user identifier of a user to be recommended; a processing module for determining, based on the user relationship graph and the user identifier, a transaction weight of the user to be recommended and a current package corresponding to the transaction weight, wherein the transaction weight includes a first weight and a second weight, the first weight indicating the importance of the current package of the user to be recommended to the user to be recommended, and the second weight indicating the importance of the current package of related users to the user to be recommended, wherein the related users are users who have a social relationship with the user to be recommended; the processing module is further configured to, based on a package recommendation model, obtain at least one package to be recommended and a recommendation weight corresponding to the package to be recommended, according to the transaction weight and the current package; the processing module is further configured to, based on the package to be recommended and the recommendation weight, determine at least one recommended package and recommend the recommended package to the user to be recommended.

[0018] According to another aspect of this disclosure, an electronic device is also provided, comprising: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to perform the package recommendation method described in any of the preceding claims by executing the executable instructions.

[0019] According to another aspect of this disclosure, a computer-readable storage medium is also provided, on which a computer program is stored, which, when executed by a processor, implements the package recommendation method described in any of the preceding claims.

[0020] According to another aspect of this disclosure, a computer program product is also provided, comprising: a computer program or instructions that, when executed by a processor, implement the package recommendation method of any of the above.

[0021] In the embodiments of this disclosure, a user relationship graph and user identifiers of users to be recommended are obtained. Based on the user relationship graph and user identifiers, the transaction weight of the user to be recommended and the current package corresponding to the transaction weight are determined. The transaction weight includes a first weight and a second weight. The first weight indicates the importance of the current package of the user to be recommended to the user, and the second weight indicates the importance of the current packages of related users to the user to be recommended. Related users are users who have social relationships with the user to be recommended. Based on the package recommendation model, at least one package to be recommended and its corresponding recommendation weight are obtained according to the transaction weight and the current package. Based on the package to be recommended and the recommendation weight, at least one recommended package is determined and recommended to the user to be recommended. This disclosure provides more personalized and reliable telecommunications service package recommendations to the user to be recommended by analyzing the current packages of the user to be recommended and its related users, and by determining recommended packages based on the package recommendation model, transaction weight, and current packages, thereby improving the accuracy of package recommendations and user satisfaction.

[0022] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0023] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure. It is obvious that the drawings described below are merely some embodiments of this disclosure, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort.

[0024] Figure 1 A schematic diagram of an exemplary application system architecture to which the package recommendation method of the present disclosure embodiments can be applied is shown.

[0025] Figure 2 A flowchart of a package recommendation method according to an embodiment of this disclosure is shown.

[0026] Figure 3 This illustration shows a user relationship graph diagram according to an embodiment of the present disclosure.

[0027] Figure 4 This illustration shows another user relationship graph diagram in an embodiment of the present disclosure.

[0028] Figure 5 This illustration shows another user relationship graph in an embodiment of the present disclosure.

[0029] Figure 6 A schematic diagram of a package recommendation device is shown in an embodiment of this disclosure.

[0030] Figure 7A structural block diagram of an electronic device according to an embodiment of the present disclosure is shown.

[0031] Figure 8 A schematic diagram of a computer-readable storage medium according to an embodiment of the present disclosure is shown. Detailed Implementation

[0032] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that this disclosure will be more comprehensive and complete, and will fully convey the concept of the exemplary embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

[0033] Furthermore, the accompanying drawings are merely illustrative of this disclosure and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0034] To facilitate understanding, before introducing the embodiments of this disclosure, the following explanations are provided for several terms involved in the embodiments of this disclosure:

[0035] Blockchain: A decentralized distributed ledger that is block-based, immutable, secure, and reliable. It combines distributed storage, peer-to-peer transmission, consensus mechanisms, cryptography, and other technologies to record transactions and information through a continuously growing chain of data blocks, ensuring data security and transparency.

[0036] Smart contracts: A set of digitally defined promises, including protocols that allow contract participants to execute these promises. A contract consists of a set of code (contract functions) and data (contract state), and runs on the Ethereum Virtual Machine.

[0037] A blockchain transaction network is a distributed network system built on blockchain technology for recording and processing various types of transactions. Blockchain is a decentralized and immutable data storage technology that allows multiple participants to transact without a central trusted party, ensuring data security and transparency.

[0038] Polkadot is a blockchain platform designed to enable cross-chain interoperability, supporting multiple independent blockchains to collaborate within the same network.

[0039] The specific implementation methods of the embodiments of this disclosure will now be described in detail with reference to the accompanying drawings.

[0040] Figure 1 A schematic diagram of an exemplary application system architecture to which the package recommendation method of the embodiments of this disclosure can be applied is shown. For example... Figure 1 As shown, the system architecture may include terminal device 101, network 102 and server 103.

[0041] Network 102 is a medium used to provide a communication link between terminal device 101 and server 103, and can be a wired network or a wireless network.

[0042] Optionally, the aforementioned wireless or wired networks use standard communication technologies and / or protocols. The network is typically the Internet, but can also be any network, including but not limited to Local Area Networks (LANs), Metropolitan Area Networks (MANs), Wide Area Networks (WANs), mobile, wired or wireless networks, private networks, or any combination of virtual private networks. In some embodiments, technologies and / or formats including Hyper Text Markup Language (HTML), Extensible Markup Language (XML), etc., are used to represent data exchanged over the network. Furthermore, conventional encryption technologies such as Secure Socket Layer (SSL), Transport Layer Security (TLS), Virtual Private Networks (VPNs), and Internet Protocol Security (IPSec) can be used to encrypt all or some links. In other embodiments, custom and / or dedicated data communication technologies can be used to replace or supplement the aforementioned data communication technologies.

[0043] Terminal device 101 can be various electronic devices, including but not limited to smartphones, tablets, laptops, desktop computers, smart speakers, smartwatches, wearable devices, augmented reality devices, virtual reality devices, etc.

[0044] Optionally, the client of the application installed on different terminal devices 101 may be the same, or the client of the same type of application based on different operating systems. Depending on the terminal platform, the specific form of the application client may also be different; for example, the application client may be a mobile client, a PC client, etc.

[0045] Server 103 can be a server that provides various services, such as a backend management server that supports the device operated by the user using terminal device 101. The backend management server can analyze and process received requests and other data, and feed the processing results back to the terminal device.

[0046] Optionally, the server can be a standalone physical server, a server cluster or distributed system consisting of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms.

[0047] Those skilled in the art will know that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative; any number of terminal devices, networks, and servers can be included depending on actual needs. This disclosure does not limit the scope of the embodiments.

[0048] Under the above system architecture, this disclosure provides a package recommendation method, which can be executed by any electronic device with computing power.

[0049] In some embodiments, the package recommendation method provided in this disclosure can be executed by a terminal device in the above-described system architecture; in other embodiments, the package recommendation method provided in this disclosure can be executed by a server in the above-described system architecture; in still other embodiments, the package recommendation method provided in this disclosure can be implemented by the terminal device and the server in the above-described system architecture through interaction.

[0050] Figure 2 This invention discloses a flowchart of a package recommendation method according to an embodiment of the present invention, as shown below. Figure 2 As shown in the embodiments of this disclosure, the package recommendation method includes the following steps.

[0051] S202, Obtain the user relationship graph and user identifiers of users to be recommended.

[0052] In this embodiment of the disclosure, the user relationship graph is used to represent the relationships between users (such as social relationships) and the relationships between users and the telecommunications products they purchase (such as NetEase Cloud Music, Youku, Tencent Meeting, etc.).

[0053] For example, such as Figure 3 As shown, User A and User B are friends, and User A and User C are spouses. User A purchased a personal communication package and online games. User B purchased a cloud conferencing enterprise VIP membership. User C purchased a yearly cloud music membership and video membership.

[0054] In an exemplary embodiment, before obtaining the user relationship graph and user identifier, the package recommendation method may further include the following steps A1 to A4.

[0055] Step A1: Obtain the user's identifier based on the account address in the blockchain transaction network.

[0056] In this embodiment of the disclosure, the user's identifier is identified through the account address in a blockchain transaction network (such as Ethereum or Polkadot). It should be noted that the user identifier can represent the user's identity or the user's contract identity.

[0057] For example, by querying a mapping table, the user identifier corresponding to the account address can be obtained. The mapping table records the correspondence between account addresses and user identifiers. For instance, account addresses a, b, and c correspond to users A, B, and C, respectively.

[0058] For example, an account address is input into a neural network model, and the output is a user identifier corresponding to the account address.

[0059] Step A2: Use knowledge graph technology to establish social relationships between user identifiers.

[0060] In this embodiment of the disclosure, knowledge graph technology is used to establish social relationships between users, such as colleagues, spouses, friends, parents, children, etc.

[0061] In this embodiment of the disclosure, such as Figure 5 As shown, for a user who wants to recommend a package (i.e., the user to be recommended), the transaction records of the user and their related social users are extracted and categorized as follows:

[0062] Contractual transactions between users and merchants (e.g., type A): Extract transaction records between users and merchants (e.g., transactions 1-5) from the blockchain, parse the smart contract content through a large model (e.g., LLM), and obtain the product type (e.g., cloud music membership, video membership).

[0063] User-to-user transfer transactions (e.g., type b): Extract the transfer amount between users (e.g., transaction 6, where the transfer amount between user A and user C accounts for 2.6%).

[0064] It should be noted that data cleaning and standardization can be performed when constructing a user relationship graph. For example, invalid transactions (such as test transactions and small, scattered transactions) can be removed. Another example is standardizing product names (such as unifying "Cloud Music Membership" as "Cloud Music Premium").

[0065] It should be noted that social relationships include transactional relationships. When a social relationship is a transactional relationship, this embodiment of the disclosure can utilize knowledge graph technology to establish transactional relationships (e.g., purchases or transfers) between user identifiers.

[0066] Step A3: Integrate user transaction data in the blockchain transaction network based on oracle and / or cross-chain bridging technology.

[0067] Step A4: Construct a user relationship graph based on social relationships and user transaction data.

[0068] In this embodiment, user transaction data from multiple blockchains (such as Ethereum and Polkadot) are integrated using oracles or cross-chain bridging technology to construct a user relationship graph. This user relationship graph is a type of cross-chain relationship graph. For example, user A's transactions on the Ethereum chain and user B's transactions on the Polkadot chain are linked through a cross-chain protocol, forming a unified social relationship network.

[0069] In this embodiment of the disclosure, the constructed user relationship graph is stored as a graph database (such as Neo4j), which supports multi-level relationship queries (such as querying the colleagues of user A's colleagues).

[0070] This disclosure identifies user identifiers through account addresses in a blockchain transaction network and integrates multi-chain transaction data to construct a cross-chain user relationship graph, effectively revealing and analyzing interactions and connections between different users. Utilizing knowledge graph technology to model social relationships between users allows for deeper understanding of users' social networks and behavioral patterns, thereby enabling more accurate personalized services, risk assessment, and data mining. Storing the user relationship graph in a graph database supports multi-level relationship queries, improving data retrieval efficiency and providing stronger support for social relationship analysis, helping all parties make more rational decisions and predictions.

[0071] S204. Based on the user relationship graph and user identifier, determine the transaction weight of the user to be recommended and the current package corresponding to the transaction weight. The transaction weight includes a first weight and a second weight. The first weight indicates the importance of the current package of the user to be recommended to the user being recommended. The second weight is used to indicate the importance of the current package of related users to the user being recommended. Related users are users who have social relationships with the user to be recommended.

[0072] In this embodiment of the disclosure, the user corresponding to the user identifier is identified as the user to be recommended, and other users related to the user to be recommended are identified as related users, wherein related users include users with multiple levels of relatedness. For example, such as Figure 4 As shown, user A is the user to be recommended, and users D and B are on the same path as user A. Users D and B are both related users of user A.

[0073] S206, based on the package recommendation model, obtains at least one package to be recommended and its corresponding recommendation weight according to the transaction weight and the current package.

[0074] In this embodiment of the disclosure, the package recommendation model is a pre-trained model. This embodiment of the disclosure does not specifically limit the type of model used for the package recommendation. For example, the package recommendation model includes a machine learning model and / or a deep learning model. For instance, the package recommendation model is a machine learning model. As another example, the package recommendation model is a deep learning model.

[0075] S208 determines at least one recommended package based on the package to be recommended and the recommendation weight, and recommends the recommended package to the user to be recommended.

[0076] This disclosure provides more personalized and reliable telecommunications service package recommendations to users by analyzing the current packages of the users to be recommended and their related users, and by determining recommended packages based on the package recommendation model, transaction weights, and current packages. This improves the accuracy of package recommendations and user satisfaction.

[0077] The present disclosure will now be described through several exemplary embodiments.

[0078] In an exemplary embodiment, the method for recommending packages provided in this disclosure, based on a user relationship graph and a user identifier, may include the following steps B1 to B4 to determine the transaction weight of the user to be recommended and the current package corresponding to the transaction weight.

[0079] Step B1: Determine the weight adjustment data based on the user relationship graph and user identifiers.

[0080] In this embodiment of the disclosure, the weight adjustment data is the data that affects the relationship weight. Adjusting the relationship weight according to the weight adjustment data can improve the accuracy of the relationship weight.

[0081] For example, weighting adjustment data may include one or more of the following: transfer amount, interaction frequency, and joint consumption records.

[0082] It should be noted that the user relationship graph is constructed based on transactions between users, and it can record data related to these transactions. For example, the user relationship graph can record data such as transfer amounts, interaction frequency, and joint consumption records.

[0083] Step B2: Obtain the initial relation weights.

[0084] In this embodiment of the disclosure, the initial relation weights can be preset values. The initial relation weights can also be set based on historical data and / or GNN training results, setting initial relation weights for different relation types.

[0085] For example, the initial relationship weight for a colleague relationship is 0.21, the initial relationship weight for a spouse relationship is 0.76, and the initial relationship weight for a friend relationship is 0.5.

[0086] It should be noted that GNN training results refer to the output obtained after training a graph neural network (GNN) model.

[0087] Step B3: Adjust the initial relation weights according to the weight adjustment data to obtain the relation weights.

[0088] For example, the weighting adjustment data includes the time interval between the most recent active periods of the relationship, the total transaction amount of the user to be recommended (e.g., the total transaction amount of the user to be recommended within 1 year), and the transaction amount between the related user and the user to be recommended.

[0089] The initial relation weights are adjusted using the following formula 1.

[0090] ω'=ω×e -λt ×(1+τ) (1)

[0091] Where ω' is the relationship weight, ω'≤1; ω is the initial relationship weight; λ is the adjustment parameter; t is the time interval between the most recent active events of the relationship; and τ is the ratio between the transaction amount between the related user and the user to be recommended and the total transaction amount of the user to be recommended.

[0092] It should be noted that the adjustment parameters can be preset values. When there is no transaction between the relevant user and the user to be recommended, the relationship weight can be the initial relationship weight.

[0093] Step B4: Based on the user relationship graph and relationship weights, obtain the transaction weight of the user to be recommended and the current package corresponding to the transaction weight.

[0094] In this embodiment of the disclosure, the current package can be the telecommunications product currently used by the user.

[0095] In one embodiment, the method for recommending service packages provided in this disclosure, based on the user relationship graph and the relationship weights, may include obtaining the transaction weight of the user to be recommended and the current service package corresponding to the transaction weight, which may include:

[0096] Based on the user relationship graph, the association path related to the user to be recommended is determined according to the preset social relationship hierarchy. The association path includes one or more of the first association path, the second association path, and the third association path. The first association path includes the user to be recommended and the current package of the user to be recommended. The second association path includes the user to be recommended, a related user, the relationship weight between the user to be recommended and the related user, and the current package of the related user. The third association path includes the user to be recommended, at least two related users, the relationship weight between the first related user and the user to be recommended, the relationship weight between two adjacent related users, and the current package of the last related user among the at least two related users.

[0097] Set the transaction weight corresponding to the current package included in the first associated path to a first value, where the first value is less than or equal to 1;

[0098] Set the transaction weight corresponding to the current package included in the second association path to the second value. The second value is equal to the relationship weight between the user to be recommended and the relevant user in the second association path. The relationship weight is less than the first value.

[0099] Set the transaction weight corresponding to the current package included in the third association path to the third value. The third value is the product of the weights of all relationships in the third association path.

[0100] The maximum value among the transaction weights corresponding to the same current package in the associated path will be used as the final transaction weight.

[0101] In this embodiment of the disclosure, the specific value of the first value can be set according to the actual application scenario and specific application experience. For example, the first value can be set to 1.

[0102] For example, such as Figure 5As shown, the social relationship level is 2. The first association path includes user A (the user to be recommended) and the user's current plan (Light Personal Communication Plan and Online Game). The transaction weight corresponding to the Light Personal Communication Plan is 1, and the transaction weight corresponding to the Online Game is 1. That is to say, the first value is 1.

[0103] It should be noted that the social relationship level is 2, which means that, starting from the user to be recommended, each association path includes a maximum of two users (one user to be recommended and one related user).

[0104] There are two second-order association paths. The first path includes User A (the user to be recommended), User B (a related user), the related user's current subscription (Cloud Meeting Enterprise VIP and Cloud Music Annual Membership), and the relationship weight between the user to be recommended and the related user (0.21). The transaction weight for Cloud Meeting Enterprise VIP and Cloud Music Annual Membership is 0.21. The second path includes User A (the user to be recommended), User C (a related user), the related user's current subscription (Video Membership and Cloud Music Annual Membership), and the relationship weight between the user to be recommended and the related user (0.76). The transaction weight for Video Membership and Cloud Music Annual Membership is 0.76. Because the transaction weight for Cloud Music Annual Membership in the second path is greater than that in the first path, the final transaction weight for Cloud Music Annual Membership is 0.76.

[0105] For example, such as Figure 4 As shown, the social relationship level is 3, and there is one third association path. The third association path includes user A (user to be recommended), user B (related user), user D (related user), the relationship weight between user A and user B, the relationship weight between user B and user D, and the current package of the last related user among at least two related users (user D's current package, i.e., Cloud Music annual membership). The transaction weight corresponding to the Cloud Music annual membership is 0.2 × 0.8 = 0.16.

[0106] In one embodiment, for a user who wants to be recommended a package (the user to be recommended), the transaction records are retrieved and the corresponding processing is performed as follows:

[0107] Direct Transaction Records (Type A): Extract your own direct contract transaction records (such as "Lightweight Personal Communication Package") and set their transaction weight to 1.

[0108] Related User Transaction Records (Category B): Transaction records of relevant users are extracted from the user relationship graph. Taking the user to be recommended as the center, the relationship weights of each associated path to that user are calculated. If the transaction involves multiple levels of association, the transaction weight of that path should be the product of the overlapping relationship weights of each node. After calculating the weights of each relationship path, the maximum value is taken as the transaction weight of that transaction record. For example, if multiple people in the relationship graph purchase the same product, the highest weight value will be used as the final transaction weight.

[0109] This embodiment of the disclosure obtains the transaction weight of the user to be recommended and the current package corresponding to the transaction weight based on the user relationship graph and relationship weight. This fully considers the influencing factors within the user's social circle, thereby recommending more accurate packages and improving user satisfaction. Furthermore, using this embodiment for package recommendation can save manpower and improve work efficiency.

[0110] In another exemplary embodiment, based on the package recommendation model, at least one package to be recommended and the recommendation weight corresponding to the package to be recommended are obtained according to the transaction weight and the current package, which may include the following steps C1 and C2.

[0111] Step C1: Obtain user characteristics, which include one or more of the following: user age, gender, duration of network access, city of residence, total historical spending, transaction frequency, average spending, and service plan preferences. It should be noted that service plan preferences can include preferences for services not currently in use (e.g., not having purchased a cloud service plan).

[0112] Step C2 involves inputting user characteristics, transaction weights, and the current package into the package recommendation model, and outputting at least one package to be recommended and its corresponding recommendation weight.

[0113] To further improve the accuracy of recommendations, the data (such as transaction weights and current packages) can be preprocessed before being input into the package recommendation model.

[0114] Transaction information and user characteristics are standardized and feature extracted. For example, for the direct transaction records of the user to be recommended (transaction weight of 1), feature extraction is performed using field extraction methods to extract the user's own purchased package name (current package), price, purchase time, and consumption amount. The package names, prices, purchase times, and adjusted transaction weights (e.g., 0.76, 0.21, etc.) of related users are also extracted. The transaction record hierarchy is labeled (Level 1 relationship: user to be recommended; Level 2 relationship: directly associated users; Level 3 relationship: associated users of associated users, and so on).

[0115] For example, the package recommendation model generates a personalized recommendation list based on the input data (user characteristics, transaction weights, and current package) and combines it with the telecommunications product database. The personalized recommendation list includes at least one package to be recommended and the recommendation weight corresponding to the package to be recommended.

[0116] This disclosure incorporates user characteristics, transaction weights, and the current service plan into a service plan recommendation model. Based on the user's personalized needs and transaction habits, it recommends the most suitable service plan (the plan to be recommended) and assigns a corresponding recommendation weight to each plan. This accurately predicts service plans that users might be interested in, thereby improving recommendation accuracy, optimizing user experience, and helping businesses increase conversion rates and user satisfaction.

[0117] In yet another exemplary embodiment, based on the package recommendation model, at least one package to be recommended and the recommendation weight corresponding to the package to be recommended are obtained according to the transaction weight and the current package, which may include the following steps D1 and D2.

[0118] Step D1: Obtain package features, which include one or more of the following: package name, package price, package validity period, data limit, additional services, and package rating.

[0119] In this embodiment of the disclosure, the supplementary service can be a service other than the main package, such as targeted data traffic or / or membership benefits. The package rating can be a matching score based on the historical behavior of the user to be recommended (e.g., if user A prefers a high data traffic package, then the high data traffic package has a higher matching score, i.e., a higher package rating).

[0120] Step D2: Input the package features, transaction weights, and current package into the package recommendation model, and output at least one package to be recommended and the recommendation weights corresponding to the package to be recommended.

[0121] This embodiment of the disclosure inputs package features, transaction weights, and the current package into the package recommendation model. By comprehensively analyzing the user's personalized needs, transaction behavior, and current package usage, it recommends the package that best meets the user's needs and calculates a corresponding recommendation weight for each package to be recommended, thereby improving the accuracy of recommendations, optimizing user experience, increasing business conversion rates, and reducing user churn.

[0122] In yet another exemplary embodiment, based on the package recommendation model, at least one package to be recommended and the recommendation weight corresponding to the package to be recommended are obtained according to the transaction weight and the current package, which may include the following steps E1 and E2.

[0123] Step E1: Obtain feature data. Feature data includes user features and package features. User features include one or more of the following: user age, gender, duration of network access, city, total historical spending, transaction frequency, average spending, and package preferences. Package features include one or more of the following: package name, package price, package validity period, data limit, additional services, and package rating.

[0124] In this embodiment of the disclosure, the supplementary service can be a service other than the main package, such as targeted data traffic or / or membership benefits. The package rating can be a matching score based on the historical behavior of the user to be recommended (e.g., if user A prefers a high data traffic package, then the high data traffic package has a higher matching score, i.e., a higher package rating).

[0125] Step E2 involves inputting the feature data, transaction weights, and current package into the package recommendation model, and outputting at least one package to be recommended and its corresponding recommendation weight.

[0126] In this embodiment, user characteristics and package characteristics data are acquired, including basic user information, consumption behavior, and detailed package attributes, laying the foundation for subsequent personalized recommendations. The characteristic data, transaction weights, and the current package are input into the package recommendation model to calculate the most suitable package for the user's needs. Each package is assigned a corresponding recommendation weight, thereby improving recommendation accuracy, optimizing user experience, and driving business conversion and user satisfaction.

[0127] In one embodiment, the package recommendation model may include a machine learning model and / or a deep learning model. For example, the package recommendation model may be a machine learning model. Or, for another example, the package recommendation model may be a deep learning model.

[0128] In this embodiment of the disclosure, the specific model used for recommending service packages can be selected based on the actual application scenario and specific application experience. For example, in the scenario of recommending telecommunications service packages, the model can be an XGBoost model, which is suitable for tabular data and can handle high-dimensional sparse features. Another example is a model based on Gradient Boosting Decision Tree (GBDT), which improves the model's generalization ability through ensemble learning.

[0129] For example, in scenarios with complex social relationships, the package recommendation model can be a graph neural network (GNN) model, where the neural network models the complex relationships between users, packages, and relationships (e.g., user A's colleague B buys package X, and user A's spouse C buys package Y).

[0130] For example, the package recommendation model could be a Transformer (a deep learning model) that can capture long sequence dependencies (such as the order in which a user has purchased packages in the past).

[0131] In this embodiment of the disclosure, the training data for training the package recommendation model may include positive samples and negative samples. For example, positive samples are packages that the user has purchased (labeled 1), and negative samples are packages that the user has not purchased (labeled 0). Positive and negative samples are sampled in a ratio (e.g., 1:1).

[0132] The training process is as follows:

[0133] Feature encoding: One-Hot encoding or embedding encoding is performed on categorical variables (such as package name, user gender).

[0134] Model training: Use cross-validation to divide the training and validation sets, and adjust hyperparameters (such as learning rate and tree depth).

[0135] Evaluation metrics: accuracy, recall, and AUC (Area Under Curve).

[0136] This disclosure does not specify how the package recommendation model should be trained.

[0137] This disclosure, through the selection of a suitable package recommendation model, can significantly improve the accuracy and effectiveness of the recommendation system. For example, the XGBoost model, by efficiently processing tabular data and high-dimensional sparse features, can provide accurate personalized recommendations in telecommunications package recommendation scenarios; while the Gradient Boosting Decision Tree (GBDT)-based model, through ensemble learning, further enhances the model's generalization ability, enabling it to provide reliable recommendation results even in changing data environments. This effectively improves the user experience and the business value of the recommendation system.

[0138] In yet another exemplary embodiment, the package recommendation method provided in this disclosure, which determines at least one recommended package based on the package to be recommended and the recommendation weight, and recommends the recommended package to the user to be recommended, includes: combining the packages to be recommended according to a preset strategy and the recommendation weight to obtain at least one recommended package.

[0139] In this embodiment of the disclosure, the preset strategy can be determined based on the current preferential policies of telecommunications products, the recommendation weight, or the nature of the package. It should be noted that a recommended package may include one or more packages to be recommended.

[0140] Based on recommendation weights and preset strategies, relevant packages are combined and bundled into new smart contracts for users to choose and use. Preset strategies may include high recommendation weight priority strategies, complementary strategies, and preferential binding strategies.

[0141] For example, the package with the highest recommendation weight (such as "China Telecom Family Broadband Package 2000M" with 76%) is used as the main package, and combined with a package with a medium recommendation weight (such as "100G of Targeted Traffic for Audio-Visual Membership" with 42%) to form a combination.

[0142] For example, based on the analysis results of the user behavior to be recommended, a package with complementary functions (such as a broadband package + cloud conferencing service) can be selected.

[0143] For example, according to the operator's policy, high-profit packages are bundled with low-profit packages for sale (such as "Cloud PC Enjoy Edition 8 Core 32G" + "Colorful Ringtone Audio and Video Ringtone VIP").

[0144] In one embodiment, a preset strategy is dynamically adjusted. For example, the discount rate (such as a "limited-time 30% off") is dynamically adjusted based on market supply and demand or inventory conditions, thereby adjusting the preset strategy. For instance, based on feedback from the behavior of users to be recommended, if the purchase rate of a certain package is lower than expected, strategy optimization is automatically triggered (such as increasing the weight of add-on packages or adjusting prices).

[0145] It should be noted that this disclosure can achieve cross-chain data synchronization. If user transaction records are distributed across multiple blockchains (such as Ethereum and Hyperledger), and smart contracts support cross-chain verification and price synchronization, the embodiments of this disclosure can ensure transaction transparency through cross-chain verification.

[0146] This disclosure embodiment combines recommended packages with preset strategies and recommendation weights to create recommended packages that better meet the needs of the users to be recommended, thereby improving the accuracy of package recommendations and user satisfaction.

[0147] In yet another exemplary embodiment, after determining at least one recommended package based on the package to be recommended and the recommendation weight, and recommending the recommended package to the user to be recommended, the package recommendation method may further include: generating a smart contract in response to the user to be recommended selecting the recommended package, so as to enable the user to be recommended to activate the package.

[0148] In this embodiment, smart contract code can be written using Solidity (a programming language for writing blockchain smart contracts), including a package description, price, and preferential terms. The referred user can then pay and activate the package with a single click using a blockchain wallet (such as MetaMask).

[0149] It should be noted that unit testing should be performed using Truffle or / and Hardhat tools to ensure the correctness of the logic (such as payment verification and refund rules).

[0150] In this embodiment of the disclosure, in response to a user's selection of a recommended package, a smart contract is generated. The user then activates the package (smart contract) with a single click via their blockchain wallet, thereby automating the transaction process. In other words, after a user selects a recommended package, this embodiment generates a corresponding smart contract and deploys it to the blockchain network. Once the user makes a one-click payment via their blockchain wallet, the smart contract automatically executes and activates the package, thus automating the transaction process.

[0151] This disclosure improves the accuracy and personalization of recommendations. By applying a package recommendation model (such as a large model) to the analysis of transaction records and social relationship graphs, it comprehensively considers the personal preferences of the user to be recommended and the behavioral patterns of their social circle (related users), generating a highly personalized list of telecommunications package recommendations. Furthermore, by combining a product database and promotional strategies, it can provide customized bundled promotional contract options for the user to be recommended, greatly enhancing the personalization of the service and the user experience. This approach surpasses traditional recommendation mechanisms that are simply based on historical consumption or basic user segmentation, ensuring that each user receives the most suitable recommendation.

[0152] This disclosure enhances data security and transparency by leveraging blockchain technology to ensure the secure storage and transmission of all transaction records while maintaining a high degree of transparency in the operational process. This not only helps protect user privacy and sensitive information from being leaked or tampered with, but also increases trust in the entire recommendation process, making users feel more confident and reliable about the recommended services. Furthermore, the automatic execution of related logic based on smart contracts further simplifies business processes and reduces the risks associated with human intervention.

[0153] The present disclosure will be further illustrated by the following examples. Assume there are three accounts (accounts a, b, and c) in a blockchain transaction network, corresponding to users A, B, and C respectively. This disclosure can be applied to smart package recommendations in a blockchain transaction network based on smart contracts.

[0154] The present disclosure provides a method for recommending a package, which includes the following steps F1 to F4.

[0155] Step F1: Obtain the user relationship graph and the user identifiers of the users to be recommended.

[0156] Before proceeding to step F1, a user relationship graph needs to be constructed. For example, blockchain identities are identified, showing that account a is bound to user A, account b to user B, account c to user C, and that user A and user B are colleagues, while user A and user C are spouses. Social relationship modeling is then performed, integrating user A's transactions on the Ethereum blockchain (e.g., purchasing a music streaming membership) with user C's transactions on the Polkadot blockchain (e.g., purchasing a video streaming membership) using cross-chain bridging technology, forming a unified cross-chain relationship graph. For instance, user B purchased a "Light Personal Communication Package" on the Ethereum blockchain, while user C purchased "100GB of dedicated data for a video streaming membership" on the Polkadot blockchain. The user relationship graph is then stored using the Neo4j graph database, supporting multi-level relationship queries. For example, the transaction records of user A's colleague B's colleague D (a 3-level relationship) can be analyzed and correlated.

[0157] Step F2: Based on the user relationship graph and user identifier, determine the transaction weight of the user to be recommended and the current package corresponding to the transaction weight.

[0158] This disclosure embodiment can adjust the initial relationship weights based on weight adjustment data. The weight adjustment data can include type A transactions (contractual transactions between users and merchants) and / or type B transactions (transfers between users). For example, transaction 1: User A purchases a "Light Personal Communication Package"; transaction 2: User A purchases "Online Games"; transaction 3: User B purchases "Cloud Meeting Enterprise VIP"; transaction 4: User C purchases a "Cloud Music Annual Membership"; transaction 5: User C purchases a "Video Membership." Transactions 1 to 5 are all type A transactions. As another example, transaction 6: User A transfers 200 yuan to User C (2.6%); transaction 7: User B transfers 50 yuan to User A (0.65%). Transactions 6 and 7 are both type B transactions.

[0159] It should be noted that the embodiments of this disclosure can use a large model (such as LLM) to parse the smart contract content and extract the product type (such as "Cloud Music Membership" or "Video Membership"). The extracted data is then cleaned. For example, product names are standardized (e.g., "Cloud Music Membership" is uniformly renamed "Cloud Music Premium").

[0160] For example, initial relationship weights: Colleague relationship (User A and B): ω = 0.212. Spouse relationship (User A and C): ω = 0.771. Adjusted relationship weights based on weight adjustment data: Colleague relationship (User A and B): ω ， =0.21. Spousal relationship (users A and C): ω ， =0.76.

[0161] For example, the weight of a transaction between users to be recommended is 1. Figure 5 As shown, User A is the user to be recommended, and the transaction weight for both "Light Personal Communication Package" and "Online Games" is 1. User B is a relevant user, and the transaction weight for both "Cloud Meeting Enterprise VIP" and "Cloud Music Membership" is 0.21. User C is a relevant user, and the transaction weight for both "Cloud Music Membership" and "Video Membership" is 0.76. It should be noted that because "Cloud Music Membership" corresponds to two transaction weights, the transaction weight of the same current package is the largest, and the transaction weight of "Cloud Music Membership" is 0.76.

[0162] Step F3: Based on the package recommendation model, a recommendation list is obtained according to the transaction weight, feature data, and the current package. The recommendation list includes at least one package to be recommended and the recommendation weight corresponding to the package to be recommended.

[0163] For example, the feature data includes user characteristics and package characteristics. User characteristics include static characteristics and dynamic characteristics. Static characteristics: User A's age = 30 years old, gender = male, city = Beijing. Dynamic characteristics: Recent 30-day spending = 500 yuan, preference for high-data packages. Package characteristics: "China Telecom Family Broadband Package 2000M": Price = 199 yuan / month, data = 2000GB. "TV & Audio Member Targeted Data 100GB": Price = 29 yuan / month, data = 100GB.

[0164] For example, the recommendation list (with recommendation weights in parentheses) is as follows:

[0165] China Telecom Family Broadband Package 2000M (76%)

[0166] China Telecom Cloud PC Enjoy Edition 8-core 32GB (71%)

[0167] 100GB (42%) of data traffic is allocated to audio-visual members.

[0168] Colorful Ringtones Audio and Video Ringtones VIP (34%).

[0169] China Telecom Cloud Meeting (12%).

[0170] Step F4: Based on the recommended service plan and the recommendation weight, determine at least one recommended service plan and recommend it to the user to be recommended.

[0171] For example, based on prioritizing higher recommendation weights, the recommended packages are determined as follows:

[0172] Main package: "China Telecom Family Broadband Package 2000M" (76%).

[0173] Additional package: "100GB of dedicated data for audio-visual members" (42%).

[0174] For example, based on the complementarity strategy, the recommended package is as follows:

[0175] China Telecom Family Broadband Package 2000M (76%) + China Telecom Cloud Meeting (12%).

[0176] For example, based on the discount bundling strategy, the recommended package is determined as follows:

[0177] China Telecom Cloud PC Enjoy Edition 8-core 32GB (71%) + Colorful Ringtones Audio and Video Ringtones VIP (34%).

[0178] Once the referred user selects a recommended package, the Solidity smart contract code is generated using a large model. The referred user then pays and activates the package through the MetaMask wallet, completing the entire referral and subscription process.

[0179] In summary, this disclosure significantly improves the matching degree between recommendation results and user needs by analyzing the transaction behavior of users and their social circles and combining it with dynamic weight adjustments. Furthermore, this disclosure enhances the adaptability and accuracy of the recommendation system by combining multi-dimensional feature data with machine learning models.

[0180] Furthermore, this disclosure enhances data security and transparency. It utilizes blockchain technology to ensure the immutability of user identities, transaction records, and recommendation logic, reducing the risk of data breaches. Smart contracts support cross-chain verification, guaranteeing transparency and trust in the transaction process.

[0181] Furthermore, this disclosure can also optimize user experience and user stickiness. Dynamically adjusting transaction weights and recommendation strategies can meet personalized user needs (such as complementary package combinations). Smart contract bundled promotions and one-click payment functions simplify the operation process and improve user satisfaction.

[0182] Furthermore, this disclosure can also promote product innovation and market responsiveness, with cross-chain data integration and real-time promotional strategies supporting operators in quickly launching innovative package combinations. Automated recommendation and contract generation mechanisms reduce operating costs and enhance market competitiveness.

[0183] Furthermore, this disclosure can simplify business processes and reduce operating costs. Smart contracts automatically execute transaction verification, payment settlement, and package activation, reducing manual intervention. Graph databases support efficient multi-level relationship queries, shortening the response time of recommendation systems.

[0184] It should be noted that the acquisition, storage, use, and processing of data in this disclosed technical solution comply with the relevant provisions of national laws and regulations. The various types of data, such as personal identity data, operational data, and behavioral data related to individuals, customers, and groups, obtained in the embodiments of this disclosure have all been authorized.

[0185] Based on the same inventive concept, this disclosure also provides a package recommendation device, as described in the following embodiments. Since the principle by which this device solves the problem is similar to that of the method embodiments described above, the implementation of this device embodiment can refer to the implementation of the method embodiments described above, and repeated details will not be repeated.

[0186] Figure 6 This diagram illustrates a package recommendation device according to an embodiment of the present disclosure, such as... Figure 6 As shown, the device includes an acquisition module 61 and a processing module 62. The acquisition module 61 can be used to acquire a user relationship graph and user identifiers of users to be recommended. The processing module 62 can be used to determine the transaction weight of the user to be recommended and the current package corresponding to the transaction weight based on the user relationship graph and user identifiers. The transaction weight includes a first weight and a second weight. The first weight indicates the importance of the current package of the user to be recommended to the user, and the second weight indicates the importance of the current package of related users to the user to be recommended. Related users are users who have a social relationship with the user to be recommended. The processing module 62 can also be used to obtain at least one package to be recommended and its corresponding recommendation weight based on a package recommendation model, according to the transaction weight and the current package. The processing module 62 can also be used to determine at least one recommended package based on the package to be recommended and its recommendation weight, and recommend the recommended package to the user to be recommended.

[0187] In one embodiment, before obtaining the user relationship graph and user identifiers, the processing module 62 can also be used to obtain the user identifiers based on the account addresses in the blockchain transaction network; establish social relationships between user identifiers using knowledge graph technology; integrate user transaction data in the blockchain transaction network based on oracle and / or cross-chain bridging technology; and construct a user relationship graph based on social relationships and user transaction data.

[0188] In one embodiment, the processing module 62 can also be used to determine weight adjustment data based on the user relationship graph and user identifier; obtain initial relationship weights; adjust the initial relationship weights based on the weight adjustment data to obtain relationship weights; and obtain the transaction weights of the users to be recommended and the current package corresponding to the transaction weights based on the user relationship graph and relationship weights.

[0189] In one embodiment, the processing module 62 can also be used to determine, based on the user relationship graph and according to a preset social relationship hierarchy, the association path related to the user to be recommended. The association path includes one or more of a first association path, a second association path, and a third association path. The first association path includes the user to be recommended and the user's current package. The second association path includes the user to be recommended, a related user, the relationship weight between the user to be recommended and the related user, and the related user's current package. The third association path includes the user to be recommended, at least two related users, the relationship weight between the first related user and the user to be recommended, the relationship weight between two adjacent related users, and the current package of the last related user among the at least two related users. The transaction weight corresponding to the current package included in the first association path is set to a first value, which is less than or equal to 1. The transaction weight corresponding to the current package included in the second association path is set to a second value, which is equal to the relationship weight between the user to be recommended and the related user in the second association path, and the relationship weight is less than the first value. The transaction weight corresponding to the current package included in the third association path is set to a third value, which is the product of all relationship weights in the third association path. The maximum value among the transaction weights corresponding to the same current package in the association path is taken as the final transaction weight.

[0190] In one embodiment, the processing module 62 can also be used to obtain user characteristics, which include one or more of the following: user age, gender, network access duration, city, total historical consumption amount, transaction frequency, average consumption amount, and package preference; input the user characteristics, transaction weights, and current package into the package recommendation model, and output at least one package to be recommended and the recommendation weight corresponding to the package to be recommended.

[0191] In one embodiment, the processing module 62 can also be used to obtain package features, which include one or more of the following: package name, package price, package validity period, data limit, additional services, and package rating; input the package features, transaction weight, and current package into the package recommendation model, and output at least one package to be recommended and the recommendation weight corresponding to the package to be recommended.

[0192] In one embodiment, the processing module 62 can also be used to acquire feature data, which includes user features and package features. User features include one or more of the following: user age, gender, network duration, city, total historical spending, transaction frequency, average spending, and package preferences. Package features include one or more of the following: package name, package price, package validity period, data limit, additional services, and package rating. The feature data, transaction weights, and the current package are input into the package recommendation model, and at least one package to be recommended and the recommendation weight corresponding to the package to be recommended are output.

[0193] In one embodiment, the package recommendation model includes a machine learning model and / or a deep learning model.

[0194] In one embodiment, the processing module 62 can also be used to combine the recommended packages according to a preset strategy and recommendation weight to obtain at least one recommended package.

[0195] In one embodiment, the processing module 62 can also be used to generate a smart contract in response to the user's selection of a recommended package, so that the user can activate the package.

[0196] The package recommendation device disclosed in this embodiment analyzes the current packages of the user to be recommended and its related users, and determines the recommended package based on the package recommendation model, transaction weight, and current package. This provides the user to be recommended with more personalized and reliable telecommunications service package recommendations, and improves the accuracy of package recommendations and user satisfaction.

[0197] It should be noted that the examples and application scenarios implemented by the modules in the above device embodiments and the corresponding steps in the method embodiments are the same, but are not limited to the content disclosed in the above method embodiments. It should also be noted that the above modules, as part of the device, can be executed in a computer system such as a set of computer-executable instructions.

[0198] Those skilled in the art will understand that various aspects of this disclosure can be implemented in the following forms: a completely hardware implementation, a completely software implementation (including firmware, microcode, etc.), or a combination of hardware and software implementations, which can be collectively referred to herein as a "circuit", "module" or "system".

[0199] Based on the same inventive concept, this disclosure also provides an electronic device, which includes: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to execute the package recommendation method of any of the above via executing the executable instructions. Since the principle by which this electronic device embodiment solves the problem is similar to that of the above method embodiments, the implementation of this electronic device embodiment can refer to the implementation of the above method embodiments, and repeated details will not be described again.

[0200] The following reference Figure 7 To describe an electronic device 700 according to such an embodiment of the present disclosure. Figure 7 The electronic device 700 shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments disclosed herein.

[0201] like Figure 7As shown, the electronic device 700 is manifested in the form of a general-purpose computing device. The components of the electronic device 700 may include, but are not limited to: at least one processing unit 710, at least one storage unit 720, and a bus 730 connecting different system components (including storage unit 720 and processing unit 710).

[0202] The storage unit stores program code that can be executed by the processing unit 710, causing the processing unit 710 to perform the steps described in the "Exemplary Methods" section above according to various exemplary embodiments of this disclosure.

[0203] Storage unit 720 may include a readable medium in the form of a volatile storage unit, such as random access memory (RAM) 7201 and / or cache memory 7202, and may further include a read-only memory (ROM) 7203.

[0204] The storage unit 720 may also include a program / utility 7204 having a set (at least one) program module 7205, such program module 7205 including but not limited to: an operating system, one or more application programs, other program modules and program data, each or some combination of these examples may include an implementation of a network environment.

[0205] Bus 730 can represent one or more of several types of bus structures, including a memory cell bus or memory cell controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus using any of the various bus structures.

[0206] Electronic device 700 can also communicate with one or more external devices 740 (e.g., keyboard, pointing device, Bluetooth device, etc.), and with one or more devices that enable a user to interact with electronic device 700, and / or with any device that enables electronic device 700 to communicate with one or more other computing devices (e.g., router, modem, etc.). This communication can be performed via input / output (I / O) interface 750. Furthermore, electronic device 700 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 760. As shown, network adapter 760 communicates with other modules of electronic device 700 via bus 730. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with electronic device 700, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0207] From the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solutions according to the embodiments of this disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, external hard drive, etc.) or on a network, including several instructions to cause a computing device (such as a personal computer, server, terminal device, or network device, etc.) to execute the methods according to the embodiments of this disclosure.

[0208] Based on the same inventive concept, in the disclosed exemplary embodiments, a computer-readable storage medium is also provided, which may be a readable signal medium or a readable storage medium. Figure 8 This illustration shows a schematic diagram of a computer-readable storage medium according to an embodiment of the present disclosure, such as... Figure 8 As shown, the computer-readable storage medium 800 stores a program product capable of implementing the methods described above.

[0209] More specific examples of computer-readable storage media in this disclosure may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0210] In this disclosure, a computer-readable storage medium may include a data signal propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A readable signal medium may also be any readable medium other than a readable storage medium, capable of transmitting, propagating, or transmitting a program for use by or in connection with an instruction execution system, apparatus, or device.

[0211] Optionally, the program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, RF, etc., or any suitable combination thereof.

[0212] In practical implementation, program code for performing the operations of this disclosure can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java and C++, and conventional procedural programming languages ​​such as C or similar languages. The program code can execute entirely on the user's computing device, partially on the user's device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0213] Based on the same inventive concept, this disclosure also provides a computer program product, including a computer program or instructions, which, when executed by a processor, implements the package recommendation method of any one of the above method embodiments. Since the principle by which this computer program product embodiment solves the problem is similar to that of the above method embodiments, the implementation of this computer program product embodiment can refer to the implementation of the above method embodiments, and repeated details will not be elaborated further.

[0214] It should be noted that although several modules or units for the device used to perform actions have been mentioned in the detailed description above, this division is not mandatory. In fact, according to embodiments of this disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0215] Furthermore, although the steps of the method in this disclosure are described in a specific order in the accompanying drawings, this does not require or imply that the steps must be performed in that specific order, or that all the steps shown must be performed to achieve the desired result. Additional or alternative steps may be omitted, multiple steps may be combined into one step, and / or a step may be broken down into multiple steps.

[0216] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This disclosure is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the appended claims.

Claims

1. A method for recommending meal packages, characterized in that, include: Obtain the user relationship graph and user identifiers of users to be recommended; Based on the user relationship graph and the user identifier, the transaction weight of the user to be recommended and the current package corresponding to the transaction weight are determined. The transaction weight includes a first weight and a second weight. The first weight indicates the importance of the current package of the user to be recommended to the user to be recommended. The second weight is used to indicate the importance of the current package of related users to the user to be recommended. The related users are users who have social relationships with the user to be recommended. Based on the package recommendation model, at least one package to be recommended and the recommendation weight corresponding to the package to be recommended are obtained according to the transaction weight and the current package. Based on the proposed package and the recommendation weight, at least one recommended package is determined and recommended to the proposed user.

2. The package recommendation method according to claim 1, characterized in that, Before obtaining the user relationship graph and user identifiers, the method further includes: Based on the account address in the blockchain transaction network, the user's identifier is obtained; Establish social relationships between the user identifiers using knowledge graph technology; Based on oracle and / or cross-chain bridging technology, integrate user transaction data in the blockchain transaction network; The user relationship graph is constructed based on the social relationships and the user transaction data.

3. The package recommendation method according to claim 1, characterized in that, Based on the user relationship graph and the user identifier, the transaction weight of the user to be recommended and the current package corresponding to the transaction weight are determined, including: Based on the user relationship graph and the user identifier, determine the weight adjustment data; Obtain the initial relation weights; Based on the weight adjustment data, the initial relationship weights are adjusted to obtain the relationship weights; Based on the user relationship graph and the relationship weights, the transaction weight of the user to be recommended and the current package corresponding to the transaction weight are obtained.

4. The package recommendation method according to claim 3, characterized in that, Based on the user relationship graph and the relationship weights, the transaction weight of the user to be recommended and the current package corresponding to the transaction weight are obtained, including: Based on the user relationship graph, a connection path related to the user to be recommended is determined according to a preset social relationship hierarchy. The connection path includes one or more of a first connection path, a second connection path, and a third connection path. The first connection path includes the user to be recommended and the user's current plan. The second connection path includes the user to be recommended, a related user, the relationship weight between the user to be recommended and the related user, and the related user's current plan. The third connection path includes the user to be recommended, at least two related users, the relationship weight between the first related user and the user to be recommended, the relationship weight between two adjacent related users, and the current plan of the last related user among the at least two related users. Set the transaction weight corresponding to the current package included in the first associated path to a first value, where the first value is less than or equal to 1; Set the transaction weight corresponding to the current package included in the second association path to a second value. The second value is equal to the relationship weight between the user to be recommended and the related user in the second association path, and the relationship weight is less than the first value. Set the transaction weight corresponding to the current package included in the third association path to a third value, wherein the third value is the product of all relationship weights in the third association path; The maximum value among the transaction weights corresponding to the same current package in the associated path is taken as the final transaction weight.

5. The package recommendation method according to claim 1, characterized in that, Based on the package recommendation model, according to the transaction weight and the current package, at least one package to be recommended and the recommendation weight corresponding to the package to be recommended are obtained, including: Acquire user characteristics, which include one or more of the following: user age, gender, network duration, city, total historical spending, transaction frequency, average spending, and package preference. The user characteristics, transaction weights, and current package are input into the package recommendation model, and at least one package to be recommended and the recommendation weight corresponding to the package to be recommended are output.

6. The package recommendation method according to claim 1, characterized in that, Based on the package recommendation model, according to the transaction weight and the current package, at least one package to be recommended and the recommendation weight corresponding to the package to be recommended are obtained, including: Obtain package features, which include one or more of the following: package name, package price, package validity period, data limit, additional services, and package rating; The package features, transaction weights, and current package are input into the package recommendation model, and at least one package to be recommended and the recommendation weight corresponding to the package to be recommended are output.

7. The package recommendation method according to claim 1, characterized in that, Based on the package recommendation model, according to the transaction weight and the current package, at least one package to be recommended and the recommendation weight corresponding to the package to be recommended are obtained, including: Acquire feature data, which includes user features and package features. The user features include one or more of the following: user age, gender, network duration, city, total historical spending, transaction frequency, average spending, and package preferences. The package features include one or more of the following: package name, package price, package validity period, data limit, additional services, and package rating. The feature data, the transaction weight, and the current package are input into the package recommendation model, and at least one package to be recommended and the recommendation weight corresponding to the package to be recommended are output.

8. The package recommendation method according to claim 1, characterized in that, The package recommendation model includes machine learning models and / or deep learning models.

9. The package recommendation method according to claim 1, characterized in that, Based on the proposed service plan and the recommendation weight, at least one recommended service plan is determined, and the recommended service plan is recommended to the proposed user, including: Based on the preset strategy and the recommendation weight, the packages to be recommended are combined to obtain at least one recommended package.

10. The package recommendation method according to claim 1, characterized in that, After determining at least one recommended package based on the package to be recommended and the recommendation weight, and recommending the recommended package to the user to be recommended, the method further includes: In response to the user's selection of the recommended package, a smart contract is generated to enable the user to activate the package.

11. A package recommendation device, characterized in that, include: The acquisition module is used to acquire user relationship graphs and user identifiers of users to be recommended; The processing module is used to determine the transaction weight of the user to be recommended and the current package corresponding to the transaction weight based on the user relationship graph and the user identifier. The transaction weight includes a first weight and a second weight. The first weight indicates the importance of the current package of the user to be recommended to the user to be recommended. The second weight is used to indicate the importance of the current package of related users to the user to be recommended. The related users are users who have social relationships with the user to be recommended. The processing module is further configured to obtain at least one package to be recommended and the recommendation weight corresponding to the package to be recommended based on the package recommendation model, according to the transaction weight and the current package; The processing module is further configured to determine at least one recommended package based on the package to be recommended and the recommendation weight, and recommend the recommended package to the user to be recommended.

12. An electronic device, characterized in that, include: processor; as well as Memory for storing the executable instructions of the processor; The processor is configured to execute the package recommendation method of any one of claims 1 to 10 by executing the executable instructions.

13. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the package recommendation method according to any one of claims 1 to 10.

14. A computer program product comprising: A computer program or instruction, characterized in that, when executed by a processor, the computer program or instruction implements the package recommendation method according to any one of claims 1 to 10.

Images