Item recommendation method and apparatus, electronic device, and storage medium

Abstract

The present disclosure provides an article recommendation method and device, electronic equipment and storage medium, relates to the field of artificial intelligence, in particular to the field of intelligent search. The specific implementation scheme is: obtaining an article scene knowledge graph constructed based on candidate articles and procurement scenes; in response to an interactive operation on a current article, selecting a target article from at least one candidate article according to the article scene knowledge graph; and outputting the target article. The technical scheme of the embodiment of the present disclosure can effectively improve the matching degree between the recommended article and the procurement party, help to reduce the procurement risk of the procurement party, and at the same time improve the cold start efficiency of the article provider.

Description

Technical Field

[0001] This disclosure relates to the field of artificial intelligence technology, and more particularly to the field of intelligent search technology, specifically to a method, apparatus, electronic device, and storage medium for recommending items. Background Technology

[0002] B2B (also written as BTB, an abbreviation for Business-to-Business) refers to a business model in which companies exchange and transmit data and information and conduct transactions through a network. It tightly integrates a company's intranet with its products and services through a B2B website or mobile client, providing better service through rapid network response and thus promoting business development. Summary of the Invention

[0003] This disclosure provides a method, apparatus, electronic device, and storage medium for recommending items.

[0004] According to one aspect of this disclosure, an item recommendation method is provided, comprising:

[0005] Obtain an item scenario knowledge graph based on candidate items and procurement scenarios;

[0006] In response to interactive operations on the current item, select the target item from at least one candidate item based on the item context knowledge graph;

[0007] Output the target item.

[0008] According to another aspect of this disclosure, an electronic device is provided, comprising:

[0009] At least one processor; and

[0010] A memory communicatively connected to the at least one processor; wherein,

[0011] The memory stores instructions that can be executed by the at least one processor, which, when executed, enables the at least one processor to perform any of the item recommendation methods provided in the embodiments of this disclosure.

[0012] According to another aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions for causing a computer to perform any of the item recommendation methods provided in the embodiments of this disclosure.

[0013] The technical solution of this disclosure improves the matching degree between the recommended items and the purchaser, reduces the purchaser's procurement risk, and improves the cold start efficiency of the item provider.

[0014] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0015] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0016] Figure 1 This is a flowchart of an item recommendation method provided according to an embodiment of the present disclosure;

[0017] Figure 2 This is a flowchart of another item recommendation method provided according to an embodiment of this disclosure;

[0018] Figure 3 This is a structural diagram of an item recommendation device provided according to an embodiment of the present disclosure;

[0019] Figure 4 This is a structural diagram of an electronic device used to implement the item recommendation method of the embodiments of this disclosure. Detailed Implementation

[0020] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0021] The item recommendation methods and devices provided in this disclosure are applicable to situations where item recommendations are made for purchasing parties. The item recommendation methods provided in the embodiments of this disclosure can be executed by an item recommendation device, which can be implemented in hardware and / or software and can be configured in an electronic device.

[0022] To facilitate understanding, this disclosure first provides a detailed explanation of the methods for recommending each item.

[0023] refer to Figure 1 The recommended methods for the items shown include:

[0024] S110. Obtain the item scenario knowledge graph constructed based on candidate items and procurement scenarios.

[0025] In this context, candidate items can be goods available for purchase by the buyer. The procurement scenario can be the application scenario of the candidate items, such as computers and keyboards used in office settings. The item scenario knowledge graph can be a knowledge graph representing the relationships between candidate items and procurement scenarios. Specifically, the item scenario knowledge graph treats each candidate item and procurement scenario as entities, and the relationships between different candidate items and procurement scenarios as the relationships between different entities.

[0026] S120. In response to the interaction with the current item, select the target item from at least one candidate item based on the item scene knowledge graph.

[0027] Interactive operations may include, but are not limited to, querying, clicking, browsing, saving, and sharing.

[0028] For example, when the purchaser performs operations such as querying, clicking, browsing, saving, or sharing the current item, the target item can be selected from the full pool of candidate items or candidate items associated with the current item, based on the item's contextual knowledge graph. The purchaser can be a unit, organization, or institution that needs to purchase the item, or the corresponding procurement equipment or procurement account.

[0029] Optionally, the target item can be selected from at least one candidate item based on a pre-trained deep learning model and an item scene knowledge graph. This disclosure does not impose any limitations on the specific network structure of the deep learning model.

[0030] For example, when a buyer needs to purchase office supplies such as computers, the buyer searches for "computer" and selects the target computer-related item from all current items based on the association between computers and office scenarios in the item scenario knowledge graph.

[0031] S130, Output target item.

[0032] For example, relevant information about the selected target item can be output to the purchasing party. This disclosure does not limit the output method of the target item; for example, it can be a display or audio output. This disclosure also does not limit the output format of the target item; for example, it can be images, text, or short videos.

[0033] The technical solution of this disclosure, based on an item-scenario knowledge graph, utilizes the association between items and scenarios to provide purchasers with target items they need to buy. By reflecting the association between items and purchase scenarios in the knowledge graph, the purchaser's item needs under different purchase scenarios are revealed, making the recommended items more closely matched to the purchase scenario. This not only efficiently and quickly recommends target items to purchasers but also improves the matching degree between the recommended target items and the purchaser, thereby reducing purchase risks and improving the cold start efficiency of item providers.

[0034] Based on the above technical solutions, this disclosure also provides an optional embodiment. In this optional embodiment, by introducing the association between items and scenarios, the matching degree between recommended items and the purchaser is improved, thereby reducing procurement risks.

[0035] See further Figure 2 The recommended methods for the items shown include:

[0036] S210. Identify candidate items that have an aggregation relationship under the same procurement scenario.

[0037] Aggregation relationships can be used to characterize the possible interactions between items in a procurement scenario. Candidate items with the same aggregation relationship indicate a higher probability that the buyer will interact with the corresponding candidate item in that procurement scenario, while candidate items without an aggregation relationship with that procurement scenario are less likely to be interacted with by the buyer in that procurement scenario.

[0038] In one optional implementation, determining candidate items with aggregation relationships under the same procurement scenario may include: obtaining a historical procurement list; determining the procurement scenario based on the item category and purchaser category of the procured items in the historical procurement list; and using the procured items in the historical procurement list under the same procurement scenario as candidate items with aggregation relationships under that procurement scenario.

[0039] The historical procurement list can be a procurement list generated by each reference purchaser in any procurement scenario. Item categories can be categorized according to their type, such as pencils belonging to stationery, mobile phones to electronic products, etc. Purchaser categories can be further categorized according to the type of purchaser, such as schools belonging to education, companies to office supplies, etc. The reference purchaser can be the same as or different from the purchaser who will subsequently perform interactive operations on the currently procured items; this disclosure does not impose any restrictions in this regard.

[0040] For example, suppose a car manufacturer's historical procurement list for its company and autonomous driving lab includes the following purchased items: LiDAR, navigation system, vehicle safety gateway, vehicle functional safety computer, antenna, laptop, pen refills, and A4 paper. Clearly, LiDAR, navigation system, vehicle safety gateway, vehicle functional safety computer, and antenna belong to autonomous driving products, corresponding to autonomous driving scenarios; laptop, pen refills, and A4 paper belong to office supplies, corresponding to office scenarios. We can consider the purchased items such as LiDAR from the historical procurement list for autonomous driving scenarios as candidate items with an aggregation relationship to that autonomous driving scenario. Similarly, we can consider the purchased items such as laptops from the historical procurement list for office scenarios as candidate items with an aggregation relationship to that office scenario.

[0041] The technical solution described above, by analyzing the procured items and procurement categories in historical procurement lists, can automatically determine the aggregation relationships of candidate items, improving the efficiency and accuracy of candidate item aggregation relationship determination. Furthermore, more accurate aggregation relationships can provide a more precise basis for constructing an item scenario knowledge graph.

[0042] S220, Treat candidate items and procurement scenarios as entities.

[0043] Here, "entity" refers to the entity in the item scenario knowledge graph. Treating candidate items and procurement scenarios as different entities lays the foundation for the subsequent construction of the item scenario knowledge graph.

[0044] S230. Construct an item scene knowledge graph based on the aggregation relationships between entities.

[0045] S240. Obtain the item scenario knowledge graph constructed based on candidate items and procurement scenarios.

[0046] S250, In response to the interaction with the current item, select the target item from at least one candidate item based on the item scene knowledge graph.

[0047] S260, Output target item.

[0048] After treating candidate items and procurement scenarios as entities in the item scenario knowledge graph, the aggregation relationship between each candidate item and each procurement scenario is treated as an entity relationship to establish the item scenario knowledge graph.

[0049] It should be noted that constructing an item scenario knowledge graph solely based on aggregation relationships cannot measure the buyer's tendency to interact with different candidate items within the same procurement scenario when using the constructed knowledge graph to determine target items. To further improve the matching degree between target items and the buyer, in an optional embodiment, quantitative data reflecting the strength of aggregation relationships can be introduced among the candidate items with aggregation relationships within the same procurement scenario during the construction of the item scenario knowledge graph. This quantitative data characterizes the buyer's tendency to interact with different candidate items within the same procurement scenario, facilitating the targeted selection and output of different candidate items within the same procurement scenario based on this quantitative data.

[0050] In one optional implementation, the step of constructing an item scenario knowledge graph based on the aggregation relationship between entities may include: determining the relationship coefficients between entities based on the interaction data generated for the aggregated candidate items in the procurement scenario; and constructing the item scenario knowledge graph based on the aggregation relationship and the relationship coefficients between entities.

[0051] Optionally, the interaction data may include relevant record data generated by the interaction behavior, such as operation records such as query, click, browse, favorite, or share. Optionally, the interaction data may also include the quantity of items purchased, i.e., the record data of the purchase amount, or other operation records, which are not limited in this embodiment.

[0052] The relationship coefficient characterizes the degree of aggregation in an aggregation relationship. A higher relationship coefficient indicates a higher degree of aggregation, meaning a stronger association between the candidate item and its corresponding procurement scenario, and a greater likelihood that the candidate item will interact with the buyer within that scenario. Conversely, a lower relationship coefficient indicates a lower degree of aggregation, meaning a weaker association between the candidate item and its corresponding procurement scenario, and a lower likelihood that the candidate item will interact with the buyer within that scenario. The technical solution described above determines the relationship coefficient between entities based on interaction data. The introduction of the relationship coefficient reflects the degree of aggregation between different candidate items and various procurement scenarios, thereby achieving differentiated aggregation of different candidate items within the same procurement scenario and laying the foundation for improving the accuracy of subsequent item recommendation results.

[0053] In one optional implementation, determining the relationship coefficient between entities based on the interaction data generated for the aggregated candidate items in the procurement scenario may include: determining the behavior coefficient between entities based on the interaction behavior data in the interaction data; determining the volume coefficient between entities based on the conversion volume data in the interaction data; and determining the relationship coefficient based on the behavior coefficient and / or the volume coefficient.

[0054] Interaction behavior data can be relevant records generated by interaction behaviors, such as records of queries, clicks, browsing, favorites, or sharing. Behavior coefficients can be determined based on interaction behavior data and are used to characterize the degree of aggregation between candidate items and procurement scenarios from the perspective of interaction behavior occurrence.

[0055] For example, at least one of the above-mentioned interactive behaviors, such as the number of times, frequency, and interaction time, can be quantified to calculate the behavior coefficient. In determining the behavior coefficient, different weights can be assigned to different interactive behaviors, and the quantified values ​​of all interactive behaviors can be weighted to obtain the final behavior coefficient. The weights assigned to the interactive behaviors can be manually set or calculated based on a large number of experiments; this embodiment does not limit this approach.

[0056] In one optional implementation, determining the behavior coefficient between entities based on the interaction behavior data in the interaction data may include: counting the number of interaction behaviors corresponding to different interaction behavior categories in the interaction behavior data; and determining the behavior coefficient based on the number of interaction behaviors corresponding to different interaction behavior categories.

[0057] The interaction behavior categories include, but are not limited to, querying, clicking, browsing, collecting, and sharing. The number of interactions corresponding to different interaction behavior categories is counted, and a behavior coefficient is calculated based on the number of interactions. For example, weights can be assigned to different interaction behavior categories, and a quantified behavior coefficient is obtained after calculation. The weights assigned to the interaction behavior categories can be set manually or calculated through extensive experimentation; this embodiment does not limit this approach.

[0058] It is understandable that the more times a certain interaction occurs, the stronger the correlation between that interaction and the procurement scenario, and the greater the weight of that interaction should be.

[0059] The technical solution described above determines behavior coefficients by statistically analyzing the number of interaction behaviors. Determining behavior coefficients based on interaction behavior types increases the richness of the criteria for determination, thereby improving the comprehensiveness of the interaction behaviors that behavior coefficients can represent. The combined determination of behavior coefficients by various interaction behavior types and their frequencies also improves the accuracy of the determination results.

[0060] Conversion volume data is used to characterize the conversion results of interactive behaviors on candidate items. For example, conversion volume data can include the purchase volume (transaction volume data) of a single purchase of a candidate item. The volume coefficient can be determined based on the conversion volume data and is used to characterize the degree of aggregation between candidate items and purchasing scenarios from the perspective of behavioral conversion.

[0061] In one optional implementation, determining the volume coefficient between entities based on the conversion volume data in the interaction data may include: determining the volume coefficient based on the ratio of the conversion volume data to the standard conversion volume data.

[0062] The standard conversion volume data can be a standard amount corresponding to the conversion volume data, representing the average conversion rate of the buyer's interaction behavior with a candidate item. It can measure the degree of conversion of interaction behavior with an item, that is, reflect the buyer's demand for the item. If the conversion volume data is greater than the standard conversion volume data, it indicates that the buyer's demand for the item is high; if the conversion volume data is much smaller than the standard conversion volume data, it indicates that the buyer's demand for the item is very low. Therefore, the ratio of the conversion volume data to the standard conversion volume data is calculated, and the ratio result is used as the volume coefficient to determine the strength of the aggregation relationship between the item and the corresponding procurement scenario. The standard conversion volume data can be determined by industry standards, can be set manually, or can be determined based on other standards. This embodiment of the disclosure does not limit this.

[0063] In a specific example, if a purchase is considered a conversion outcome, then the purchase volume can be used as conversion volume data. Correspondingly, the standard conversion volume data can be the standard purchase volume, which measures the amount of purchases. The magnitude of this volume coefficient is represented by the ratio between the purchase volume and the standard purchase volume.

[0064] For example, suppose a buyer's historical purchase records show a single purchase quantity of 1000 for a certain item, with a standard conversion volume of 500. The volume coefficient would be 2, indicating a high demand for the item and a strong aggregation relationship between the item and the purchase scenario. Conversely, if a single purchase quantity is only 50, and the standard conversion volume is 500, the volume coefficient would be 0.1, indicating a lower demand for the item and a weaker aggregation relationship between the item and the purchase scenario.

[0065] The technical solution of the above implementation method determines the volume coefficient based on the ratio of the conversion volume data to the standard conversion volume data, providing an effective method for calculating the volume coefficient. It can accurately obtain the volume coefficient that reflects the aggregation intensity of the conversion volume data and the procurement scenario, providing a foundation for establishing an item scenario knowledge graph, which helps to further improve the matching degree between items and procurement parties, thereby reducing procurement risks.

[0066] In a specific example, the purchase volume of different candidate items can be quantified to calculate the volume coefficient of that candidate item. Understandably, a larger purchase volume indicates a stronger demand for that candidate item in the corresponding procurement scenario, and therefore the volume coefficient should also be larger.

[0067] It should be noted that the relationship coefficient can be determined directly from the behavioral coefficient alone (the behavioral coefficient can be directly used as the relationship coefficient); it can also be determined directly from the volume coefficient alone (similarly, the volume coefficient can be directly used as the relationship coefficient); or weights can be assigned to both the behavioral coefficient and the volume coefficient, and the result of the weighted calculation can be used as the relationship coefficient. The weights assigned to the behavioral coefficient and the volume coefficient can be set manually or calculated based on a large number of experiments; this embodiment does not limit this.

[0068] For example, when selecting target items using an item scenario indication map containing relationship coefficients, the item recommendation probability of each candidate item in different procurement scenarios can be determined based on the relationship coefficients; by statistically analyzing the item recommendation probabilities and values ​​of candidate items in each procurement scenario, the scenario recommendation probability in the corresponding procurement scenario can be obtained; the procurement scenario with a higher scenario recommendation probability can be selected as the target scenario; and the candidate items with a higher item recommendation probability in the target scenario can be selected as the target items.

[0069] For example, in the item scenario knowledge graph, candidate items with aggregation relationships in Scenario I include items A, B, C, and D; and candidate items with aggregation relationships in Scenario II include items A, H, and K. In response to an interaction with a current item, the recommendation probability of item A in Scenario I is determined to be 0.48, and in Scenario II, it is 0.69; the recommendation probability of item C in Scenario I is 0.98; and the recommendation probability of item K in Scenario II is 0.3. Calculations show that the scenario recommendation probability in Scenario I is 0.48 + 0.98 = 1.46, and the scenario recommendation probability in Scenario II is 0.69 + 0.3 = 0.99. Since 1.46 > 0.99, Scenario I is selected as the target scenario, and the item with the highest recommendation probability in Scenario I (e.g., item C with the highest recommendation probability) is selected as the target item.

[0070] In the above implementation, the relationship coefficient is determined based on the behavior coefficient and / or the size coefficient, which improves the relationship coefficient determination mechanism and enhances the richness and comprehensiveness of the relationship coefficient determination results. This allows for the characterization of the degree of aggregation between candidate items and procurement scenarios from different dimensions, thereby improving the accuracy of the relationship coefficient determination results. This provides a foundation for more accurately establishing an item scenario knowledge graph, which helps to improve the matching degree between items and procurement parties, and thus reduces procurement risks.

[0071] The technical solution of this disclosure identifies candidate items with the same aggregation relationship as the procurement scenario, and uses the candidate items and the procurement scenario as entities in the item scenario knowledge graph to be constructed. Based on the aggregation relationship between the entities, the item scenario knowledge graph is finally constructed. The advantage of this approach is that by constructing the item scenario knowledge graph based on the aggregation relationship between candidate items and the procurement scenario, the knowledge graph can accurately reflect the relevant information between each candidate item and the procurement scenario. This allows for more precise item recommendations when the buyer is making a purchase, better matching the buyer's required procurement scenario and candidate items, thereby improving the matching degree between items and the buyer, and ultimately reducing procurement risk.

[0072] As an implementation of the above-described item recommendation methods, this disclosure also provides an optional embodiment of an execution device for implementing the above-described item recommendation methods.

[0073] See further Figure 3 The item recommendation device 300 specifically includes: a knowledge graph acquisition module 310, a target item selection module 320, and a target item output module 330, wherein...

[0074] The knowledge graph acquisition module 310 is used to acquire the item scenario knowledge graph constructed based on candidate items and procurement scenarios;

[0075] The target item selection module 320 is used to select a target item from at least one candidate item in response to the interaction operation on the current item, based on the item scene knowledge graph.

[0076] The target item output module 330 is used to output the target item.

[0077] The technical solution of this disclosure, based on an item scenario knowledge graph, utilizes the association between items and procurement scenarios to provide the purchaser with target items to buy. By reflecting the association between items and procurement scenarios in the knowledge graph, the purchaser's item needs under different procurement scenarios are revealed, making the recommended items more closely matched to the procurement scenarios. This not only efficiently and quickly recommends target items to the purchaser but also improves the matching degree between the recommended target items and the purchaser, thereby reducing procurement risks and improving the cold start efficiency of the item provider.

[0078] In one optional implementation, the knowledge graph acquisition module 310 may include:

[0079] The candidate item determination unit is used to determine candidate items that have an aggregation relationship under the same procurement scenario;

[0080] The entity determination unit is used to treat candidate items and procurement scenarios as entities;

[0081] The knowledge graph construction unit is used by the knowledge graph construction unit to construct an item scene knowledge graph based on the aggregation relationship between entities.

[0082] In one optional implementation, the candidate item determination unit may include:

[0083] The Historical Procurement List Acquisition sub-unit is used to retrieve historical procurement lists;

[0084] The procurement scenario determination sub-unit is used to determine the procurement scenario based on the item category and purchaser category of the items procured in the historical procurement list;

[0085] The candidate item determination subunit is used to select purchased items from the historical procurement list under the same procurement scenario as candidate items with aggregation relationship under that procurement scenario.

[0086] In one alternative implementation, the knowledge graph construction unit may include:

[0087] The relation coefficient determination subunit is used to determine the relation coefficients between entities based on the interaction data generated for the aggregated candidate items in the procurement scenario.

[0088] The knowledge graph construction subunit is used to construct an item scene knowledge graph based on the aggregation relationship and relationship coefficient between entities.

[0089] In one alternative implementation, the relation coefficient determination subunit may include:

[0090] The behavior coefficient determination unit is used to determine the behavior coefficients between entities based on the interaction behavior data in the interaction data;

[0091] The volume coefficient is determined from the unit and is used to determine the volume coefficient between entities based on the transformed volume data in the interaction data;

[0092] Relationship coefficients are determined from the unit and are used to determine relationship coefficients based on behavior coefficients and / or size coefficients.

[0093] In one alternative implementation, the behavior coefficient determination from the unit may include:

[0094] The interaction behavior statistics sub-unit is used to count the number of interaction behaviors corresponding to different interaction behavior categories in the interaction behavior data;

[0095] The behavior coefficient determination sub-unit is used to determine the behavior coefficient based on the number of interaction behaviors corresponding to different interaction behavior categories.

[0096] In one alternative implementation, the volume coefficient determination from the unit may include:

[0097] The volume coefficient determination sub-unit is used to determine the volume coefficient based on the ratio of the converted volume data to the standard converted volume data.

[0098] The above-mentioned products can perform the item recommendation method provided in any embodiment of this disclosure, and have the corresponding functional modules and beneficial effects for performing each item recommendation method.

[0099] The technical solutions disclosed herein, including the collection, storage, use, processing, transmission, provision, and disclosure of item scene knowledge graphs, interactive operations, and interactive data, all comply with relevant laws and regulations and do not violate public order and good morals.

[0100] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0101] Figure 4 A schematic block diagram of an example electronic device 400 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0102] like Figure 4 As shown, device 400 includes a computing unit 401, which can perform various appropriate actions and processes based on a computer program stored in read-only memory (ROM) 402 or a computer program loaded from storage unit 408 into random access memory (RAM) 403. RAM 403 may also store various programs and data required for the operation of device 400. The computing unit 401, ROM 402, and RAM 403 are interconnected via bus 404. Input / output (I / O) interface 405 is also connected to bus 404.

[0103] Multiple components in device 400 are connected to I / O interface 405, including: input unit 406, such as keyboard, mouse, etc.; output unit 407, such as various types of monitors, speakers, etc.; storage unit 408, such as disk, optical disk, etc.; and communication unit 409, such as network card, modem, wireless transceiver, etc. Communication unit 409 allows device 400 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0104] The computing unit 401 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 401 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 401 performs the various methods and processes described above, such as the item recommendation method. For example, in some embodiments, the item recommendation method may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 408. In some embodiments, part or all of the computer program may be loaded and / or installed on device 400 via ROM 402 and / or communication unit 409. When the computer program is loaded into RAM 403 and executed by the computing unit 401, one or more steps of the item recommendation method described above may be performed. Alternatively, in other embodiments, the computing unit 401 may be configured to perform the item recommendation method by any other suitable means (e.g., by means of firmware).

[0105] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0106] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0107] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0108] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0109] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with embodiments of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0110] Computer systems can include clients and servers. Clients and servers are generally geographically separated and typically interact via communication networks. The client-server relationship is established by computer programs running on the respective computers and having a client-server relationship with each other. A server can be a cloud server, also known as a cloud computing server or cloud host, a hosting product within the cloud computing service ecosystem that addresses the management difficulties and weak business scalability inherent in traditional physical hosting and VPS services. Servers can also be servers for distributed systems or servers integrated with blockchain technology.

[0111] Artificial intelligence (AI) is the study of enabling computers to simulate certain human thought processes and intelligent behaviors (such as learning, reasoning, thinking, and planning). It encompasses both hardware and software technologies. AI hardware technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, and big data processing. AI software technologies mainly include computer vision, speech recognition, natural language processing, machine learning / deep learning, big data processing, and knowledge graph technologies.

[0112] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution provided in this disclosure can be achieved, and this is not limited herein.

[0113] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A method for recommending items, including: Obtain an item scenario knowledge graph based on candidate items and procurement scenarios; In response to an interaction with the current item, a target item is selected from at least one of the candidate items based on the item scene knowledge graph. Output the target item; The item scene knowledge graph is constructed based on the following method: Obtain historical purchase lists; The procurement scenario is determined based on the item category and purchaser category of the purchased items in the historical procurement list; The purchased items in the historical purchase list under the same purchase scenario are regarded as candidate items with an aggregation relationship under that purchase scenario; wherein, the aggregation relationship is used to characterize the possible situations in which items can be interacted under the purchase scenario; Treat the candidate items and the procurement scenario as entities; Based on the interaction behavior data generated by the aggregated candidate items in the procurement scenario, determine the behavior coefficients between entities; Based on the conversion volume data in the interaction data, determine the volume coefficient between entities; The relationship coefficient is determined based on the behavior coefficient and / or the size coefficient; Based on the aggregation relationships and relationship coefficients between entities, construct the item scene knowledge graph.

2. The method of claim 1, wherein, The step of determining the behavioral coefficients between entities based on the interaction behavior data in the interaction data includes: Count the number of interaction behaviors corresponding to different interaction behavior categories in the interaction behavior data; The behavior coefficient is determined based on the number of interaction behaviors corresponding to different interaction behavior categories.

3. The method according to claim 1, wherein, The step of determining the volume coefficient between entities based on the conversion volume data in the interaction data includes: The volume coefficient is determined based on the ratio of the converted volume data to the standard converted volume data.

4. An item recommendation device, comprising: The knowledge graph acquisition module is used to acquire an item scenario knowledge graph constructed based on candidate items and procurement scenarios; The target item selection module is used to select a target item from at least one candidate item in response to an interactive operation on the current item, based on the item scene knowledge graph. The target item output module is used to output the target item; The knowledge graph acquisition module includes: The candidate item determination unit is used to determine candidate items that have an aggregation relationship under the same procurement scenario; An entity determination unit is used to identify the candidate items and the procurement scenario as entities; A knowledge graph construction unit is used to construct the item scene knowledge graph based on the aggregation relationship between entities. The knowledge graph construction unit includes: The relationship coefficient determination subunit is used to determine the relationship coefficients between entities based on the interaction data generated for the aggregated candidate items in the procurement scenario. The knowledge graph construction subunit is used to construct the item scene knowledge graph based on the aggregation relationship and the relationship coefficient between entities; The candidate item determination unit includes: The Historical Procurement List Acquisition sub-unit is used to retrieve historical procurement lists; The procurement scenario determination subunit is used to determine the procurement scenario based on the item category and purchaser category of the purchased items in the historical procurement list; The candidate item determination subunit is used to select the purchased items in the historical purchase list under the same purchase scenario as candidate items with an aggregation relationship under that purchase scenario; wherein, the aggregation relationship is used to characterize the possible situations in which items can be interacted under the purchase scenario; The relationship coefficient determination subunit includes: The behavior coefficient determination unit is used to determine the behavior coefficients between entities based on the interaction behavior data in the interaction data; The volume coefficient determination unit is used to determine the volume coefficient between entities based on the transformed volume data in the interaction data; The relation coefficient determination unit is used to determine the relation coefficient based on the behavior coefficient and / or the volume coefficient.

5. The apparatus according to claim 4, wherein, The behavior coefficients are determined from the unit, including: The interaction behavior statistics sub-unit is used to count the number of interaction behaviors corresponding to different interaction behavior categories in the interaction behavior data; The behavior coefficient determination sub-unit is used to determine the behavior coefficient based on the number of interaction behaviors corresponding to different interaction behavior categories.

6. The apparatus according to claim 4, wherein, The volume coefficient is determined from the unit, including: The volume coefficient determination sub-unit is used to determine the volume coefficient based on the ratio of the converted volume data to the standard converted volume data.

7. An electronic device, comprising: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-3.

8. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-3.

9. A computer program product comprising a computer program / instructions that, when executed by a processor, implement the steps of the method according to any one of claims 1-3.

Images