Method, device, apparatus and program product for recommending articles

Abstract

The present disclosure relates to a method, an apparatus, a device, and a program product for article recommendation. The method comprises detecting a life cycle, wherein the life cycle indicates how a target user's degree of interest in an article changes over time. The method further comprises generating user characteristics of the target user based on the life cycle. Furthermore, the method comprises generating recommendation information based on the user characteristics in order to provide the target user with an article recommendation. In this way, the life cycle is taken into account when determining the user characteristics and recommendation information is generated based thereon, whereby the required article can be recommended to the user at the appropriate time, thus improving the user experience.

Description

FIELD OF THE INVENTION

[0001] The present disclosure relates to the field of Internet technology, in particular to a method, apparatus, device and program product for recommending articles. STATE OF THE ART

[0002] With the rapid development of Internet technology, the number of users is growing rapidly, and the types and quantities of items such as products and services provided over the Internet are also steadily increasing. Since the vast amount of information on items requires users to spend a lot of time finding the desired item, the user experience of the search process is affected by the amount of irrelevant information. To solve this problem, personalized recommendation technology has increasingly become a research focus. Personalized recommendation technology can use technologies such as machine learning to analyze users' behavior, personal preferences, and other information data to recommend items they are interested in, thus improving the user experience and optimizing resource allocation. DISCLOSURE OF THE INVENTION

[0003] Embodiments of the present disclosure provide a method, apparatus, device, and program product for article recommendation. According to a first aspect of the present disclosure, a method for recommending articles is provided. The method includes detecting a life cycle, wherein the life cycle indicates how a target user's level of interest in an article changes over time. The method further includes generating user characteristics of the target user based on the life cycle. Furthermore, the method includes generating recommendation information based on the user characteristics to provide the target user with an article recommendation.

[0004] According to a second aspect of the present disclosure, a device is provided. The device comprises a lifecycle acquisition module configured to acquire a lifecycle, wherein the lifecycle indicates how the target user's degree of interest in an item changes over time. The device further comprises a feature generation module configured to generate the target user's user characteristics based on the lifecycle. Furthermore, the device comprises a recommendation module configured to generate recommendation information based on the user characteristics to provide the target user with an item recommendation.

[0005] According to a third aspect of the present disclosure, an electronic device is provided. The electronic device comprises one or more processors; and a memory device for storing one or more programs, wherein one or more of the programs, when executed by the one or more processors, cause the processor(s) to perform a method according to the first aspect of the present disclosure.

[0006] According to a fourth aspect of the present disclosure, a computer-readable storage medium is provided. Computer-executable instructions are stored on this computer-readable storage medium, wherein the computer-executable instructions are executed by a processor to implement the method provided in the first aspect according to the present disclosure.

[0007] According to a fifth aspect of the present disclosure, there is provided a computer program product physically stored on a non-transitory, computer-readable medium and comprising machine-executable instructions, the machine-executable instructions, when executed, causing the machine to perform the method provided in the first aspect of the present disclosure.

[0008] It should be understood that the content described in the disclosure of the invention is neither intended to define the essential or important features of the embodiments of the present disclosure nor to limit the scope of the present disclosure. Other features of the present disclosure will be readily understood from the following description. DESCRIPTION OF THE CHARACTERS

[0009] The above and other features, advantages, and aspects of various embodiments of the present disclosure will become more apparent from the following detailed description taken in conjunction with the figures. In the figures, like or similar reference numerals represent like or similar elements, where: Fig. 1 shows a schematic diagram of an example environment in which several embodiments of the present disclosure may be implemented; Fig. 2 shows a flowchart of a method for article recommendation according to some embodiments of the present disclosure; Fig. 3 shows a schematic representation of an example scenario for article recommendation to a user according to some embodiments of the present disclosure; Fig.4 shows a schematic representation of a life cycle determination scenario according to some embodiments of the present disclosure; Fig. 5 shows a block diagram of an article recommendation device according to some embodiments of the present disclosure; and Fig. 6 shows a block diagram of a device that can implement several embodiments of the present disclosure. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0010] It is understood that the data relating to the technical solutions described in this disclosure (including, but not limited to, data and its collection or use) must comply with the requirements of the relevant legal provisions and regulations. It is understood that the data relating to user interactions or interactions between users and content or products, as well as data related to user actions (including, but not limited to, data for analysis, storage, display, etc.)) may only be recorded, collected, or stored with the user's consent or after comprehensive authorization of all parties involved. Furthermore, the collection, use, and processing of the data in question must comply with the relevant laws, regulations, and standards of the respective countries and regions, and appropriate access options must be provided so that the user can grant or refuse authorization. It is understood that the collection, storage, use, processing, transmission, provision, and publication of user-related information in the technical solutions of the embodiments of the present disclosure comply with the relevant legal regulations and do not violate public order and morality.

[0011] In the following, the embodiments of the present disclosure are described in more detail with reference to the figures. While some embodiments of the present disclosure are illustrated in the figures, it is to be understood that the present disclosure may be embodied in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a thorough and complete understanding of the present disclosure. It should be understood that the figures and embodiments of the present disclosure serve only as examples and are not intended to limit the scope of the disclosure.

[0012] In describing the embodiments of the present disclosure, the term "comprising" and similar terms should be understood as openly encompassing, that is, "including, but not limited to." The term "based on" should be understood as "based at least in part on." The terms "an embodiment" or "this embodiment" should be understood as "at least one embodiment." The terms "first," "second," etc., may refer to different or the same items. The following may contain additional explicit and implicit definitions.

[0013] As previously described, recommending items that users are interested in can improve the user experience. In some cases, a user's interest in an item may vary over time, and in some cases, the items they prefer may be different across different time periods. For example, a user may only be interested in a particular item during a certain period of time and, after that period, lose interest in it and move on to another item. If the recommended items are not adjusted to these temporal changes in user interest, it is possible that users may be recommended items they are not interested in during certain time periods. This would not only not meet their needs but also negatively impact their user experience.

[0014] Therefore, in the embodiments of the present disclosure, a method for article recommendation is presented. In the embodiments of the present disclosure, a life cycle indicating how the target user's interest in an article changes over time can be detected, and based on this life cycle, the target user's user characteristics can be generated. Based on these user characteristics, recommendation information can be generated to recommend articles to the target user.

[0015] In this way, the generated user characteristics can be linked to the lifecycle of the user's interest in an item, ensuring that the generated recommendation information matches the temporal changes in user interest in items, so that the recommended items meet the user's needs. The embodiments of the present disclosure allow suitable items to be recommended to the user during the periods in which they need an item, thereby improving the user experience.

[0016] Fig. 1 shows a schematic diagram of an example environment 100 in which several embodiments of the present disclosure may be implemented. As in Fig.1, the environment 100 comprises a user 101, a user device 102, a network 103, and a server 104. The user device 102 may be a device with a screen and / or speakers and includes, but is not limited to, mobile phones, tablets, notebooks, portable computers, desktop computers, wearable devices, virtual reality devices, augmented reality devices, or smart home devices. The network 103 may be a medium that provides a communication connection between the user device 102 and the server 104 and may include, among other things, wired or wireless communication connections or a fiber optic cable connection. The server 104 may be a server that provides various services, for example, a backend management server that supports the operations performed for the user device 102.

[0017] In environment 100, user 101 can retrieve information via user device 102. User device 102 can receive the user data entered by user 101 and receive data and information from server 104 via network 103 and display them to user 101. Server 104 can receive data and information from user device 102 and send data and information to user device 102. In some embodiments, server 104 can collect user data from user device 102, generate user characteristics of user 101 based thereon, and generate recommendation information for article recommendation based on the user characteristics. The user device can retrieve the recommendation information generated by server 104 via network 103 and display the articles specified in the recommendation information to user 101.

[0018] In the embodiments of the present disclosure, an item may be a product offered to the user for purchase or use, including, but not limited to, electronic devices, household appliances, furniture, stationery, or vehicles, and may be durable or non-durable goods, with the embodiments of the present disclosure not imposing any limitations in this regard. In some embodiments, the item suggestion may be a product category; in other embodiments, the item may be a specific product. For example, in some embodiments, the object recommended in the recommendation information may be the product category "motor vehicle," while in other embodiments, the item recommended in the recommendation information may be a motor vehicle of a specific model.

[0019] It should be understood that environment 100 is merely one example of an embodiment of the present disclosure and does not limit the disclosure. In some embodiments, environment 100 may include multiple user devices 102, where server 104 may send recommendation information to multiple user devices 102 so that the multiple user devices can display object-related information to user 101. In some embodiments, server 104 included in environment 100 may also include one or more other devices that may perform the server functions either independently or jointly.

[0020] Fig.2 shows a flowchart of a method 200 for article recommendation according to some embodiments of the present disclosure. The method 200 can be executed by a recommendation device, wherein the recommendation device can be, for example, the server 104 in the environment 100. In the following, the method 200 is explained by way of example under the assumption that the recommendation device serves as the executing unit. With reference to Fig. 2, the method 200 may include block 202 to block 206. In block 202, the recommender captures the lifecycle, which indicates how the target user's interest in an item changes over time.

[0021] The level of interest is related to the user's need for an item, and the lifecycle can represent the change in the user's level of interest in an item over time, meaning that the lifecycle can be used to determine the user's need for an item over a specific period of time. In some embodiments, the lifecycle can include multiple lifecycle phases, with each lifecycle phase corresponding to a different level of interest. For example, the user's lifecycle for an item can include four lifecycle phases: "awareness," "liking," "purchase," and "resting," with each lifecycle phase corresponding to a specific level of interest. The lifecycle phase can further indicate the period of time during which that level of interest exists.It should be understood that the examples of life cycle phases provided herein are for illustrative purposes only and do not limit the embodiments of the present disclosure. In the embodiments of the present disclosure, the life cycle phases included in the life cycle may also have other predefined numbers and types.

[0022] In some embodiments, the recommender may generate the target user's lifecycle for an item based on the user data associated with the target user, and in some embodiments, the recommender may retrieve the target user's lifecycle for an item from other devices used to generate or store the lifecycle. In some embodiments, the lifecycle captured by the recommender may include only one or more lifecycle phases of the target user's complete lifecycle for an item. In some embodiments, the lifecycle captured by the recommender may include all lifecycle phases of the target user's complete lifecycle for an item.

[0023] In block 204, the recommender generates the user characteristics of the target user based on the life cycle. The user characteristics of the target user may indicate basic user characteristics such as age information, gender information, language information, education information, occupation information, income information, or location information. In embodiments of the present disclosure, the user characteristics of the target user generated by the recommender may also indicate the life cycle of the target user for an item. In some embodiments, to simplify data processing, the recommender may generate an embedding based on the basic user characteristics of the target user and the life cycle of the target user for an item, which serves as the user characteristics of the target user.The embedded user is a relatively low-dimensional representation of a higher-dimensional vector derived from the user characteristics and transformed. By converting various user characteristics into embedded users, these characteristics can be merged, making data processing more efficient.

[0024] In block 206, the recommendation device generates recommendation information for the target user based on the user characteristics. Based on the user characteristics, the recommendation device can determine the item most needed by the target user from the plurality of items, and the recommendation device can generate the recommendation information for this item. In some embodiments, the recommendation device can capture the characteristics of multiple items and, by matching the features, determine the item from the plurality of items to recommend to the target user. In some embodiments, the recommendation device can determine the recommendation information based on the user characteristics and a predefined recommendation model.

[0025] In some embodiments, the recommendation device may generate the recommendation information at a time appropriate for an item according to the target user's life cycle. For example, the recommendation device may generate recommendation information for recommending an item when the target user's current life cycle phase is in a predefined life cycle phase. In some embodiments, the recommendation information generated by the recommendation device may correspond to the different life cycle phases in the target user's life cycle and indicate which item is recommended to the target user in each of the different life cycle phases in which the target user is located.

[0026] In this way, the specific user characteristics can indicate the user's life cycle for an item, so that the items recommended to the user are tailored to the user's life cycle for the respective item, so that the generated recommendation information meets the user's needs for the item during different periods. This can ensure that the item recommendation meets the user's needs and recommends suitable items to the user at the right time, thereby improving the user experience.

[0027] In some embodiments, the recommender may collect data sets including user data and user-item interaction data and, based on this, determine the target user's lifecycle for the item and then recommend items to the target user. Fig.3 shows a schematic representation of an example scenario 300 for article recommendation for a user in some embodiments of the present disclosure. Fig. The scenario 300 illustrated in Figure 3 includes a data source 310, a recommendation device 320, and a user device 330. Recommendation device 320 may, for example, correspond to server 104 in environment 100, and user device 330 may, for example, correspond to user device 102 in environment 100.

[0028] The data source 310 may include a user database 311, an interaction database 312, and an item database 313. The user database 311 may store the user data of multiple users. User data may indicate information such as the user's age, gender, language, educational background, occupation, income, and residential address. The interaction database 312 may store interaction data between the user and the item, and the interaction data may indicate information about the user's interaction with the item, including, but not limited to, the number of times the user views an item, the number of searches for the item, the number of clicks on the item, the addition of the item to the shopping cart, the purchase of the item, and the user's ratings for the item.The item database 313 may store item data indicating the characteristics of an item, including, but not limited to, attributes, categories, and pricing of the item.

[0029] It should be understood that the user database 311 may contain multiple user data associated with different users, including the user data corresponding to the target user. The interaction database 312 may store multiple interaction data between a user and multiple items, as well as multiple interaction data between multiple users and an item, including the interaction data corresponding to the target user. The item database 313 may contain multiple item data corresponding to multiple items.

[0030] The recommendation device 320 may include a lifecycle detection module 321, a feature generation module 322, and a recommendation module 323. The lifecycle detection module 321 may retrieve interaction data from the interaction database 312 and, based on this, determine the lifecycle of the target user for the item. In some embodiments, the lifecycle detection module 321 may determine the interaction type and / or the number of interactions between the target user and the item based on the interaction data and, based on this, may determine the lifecycle phase of the target user for the item using predefined lifecycle determination rules, thereby determining the lifecycle of the target user for the item. The lifecycle determination rules may, for example, be a mapping between the number of interactions within a predefined period and a lifecycle phase.

[0031] As an example, the lifecycle may include the seven lifecycle stages of "awareness," "interest," "consideration," "purchase," "reinforcement," "super-member," and "reactivation." The "awareness" stage indicates that the user is beginning to learn about the item, the "interest" stage indicates that the user has increasing interest in the item, the "consideration" stage indicates that the user is evaluating and comparing the item, the "purchase" stage indicates that the user is acquiring the item, the "reinforcement" stage indicates that the user continues to support the item, the "super-member" stage indicates that the user is loyal to the item, and the "activation pending" stage indicates that the user is not interested in the item. For example, the predefined lifecycle determination rules may be as shown in Table 1: Table 1 Interaction between user and article Life cycle phase Within 30 days, views greater than or equal to 1 and less than 5 perception Within 30 days, views greater than or equal to 5 and less than 10 interest Within 30 days, views greater than 10 or Add to cart Consideration Within 30 days 1 purchase purchase Within 30 days 1 purchase and then greater than or equal to 5 views Reinforcement More than 1 purchase within 60 days Premium Member No views within 30 days Activation pending

[0032] It should be understood that the lifecycle stages and lifecycle determination rules presented above are merely examples of embodiments of the present disclosure and should not be considered as limiting these embodiments. In some embodiments, the user's lifecycle for the item may also be determined based on other rules. In some embodiments, the lifecycle stages included in the lifecycle may also have different numbers and types. In some embodiments, the lifecycle stages included in the lifecycle may be predefined or user-configured, and the rules for determining the lifecycle stages based on the interaction data may also be predefined or user-configured.

[0033] The feature generation module 322 can retrieve user data associated with the target user in the user database 311, item data associated with multiple items in the item database 313, and the target user's lifecycle for the item determined by the lifecycle acquisition module 321. The feature generation module 322 can generate item features based on the item data and user features based on the user data and the lifecycle. The recommendation module 323 can determine recommendation information for item recommendation to the target user based on the user features and the item features. Based on this, the recommendation device 320 can send the recommendation information to the user device 330, so that the user device 330 presents item-relevant information to the target user based on the recommendation information, thus implementing the item recommendation for the target user.

[0034] It should be understood that the Fig.3 is merely a schematic representation of one embodiment of the present disclosure and should not be considered limiting of the disclosure. In some embodiments, the data source 310 may correspond to multiple devices or platforms, and the recommender 320 may retrieve data from multiple devices or platforms. In some embodiments, the data in the various databases of the data source 310 may be generated by the recommender, and the recommender 320 may retrieve data associated with users and items from multiple devices or platforms, filter and integrate this data to obtain user data, interaction data, and item data, and store them in the user database 311, the interaction database 312, and the item database 313, respectively.In some embodiments, the various databases of data source 310 may each be independent stores, while in some embodiments, the various databases of data source 310 may be combined into a single store, which store may, for example, be combined with the recommendation engine.

[0035] In some embodiments, the recommender may use the interaction data to determine the interaction information that has already occurred between the target user and a specific item (the so-called target item) before the current time. Based on this, the recommender may determine the life cycle between the target user and the target item before the current time. Based on this life cycle, the recommender may generate the life cycle between the target user and the target item after the current time, which life cycle indicates the target user's degree of interest in the target item after the current time. Based on this, the recommender may generate user characteristics and generate recommendation information for recommending the item to the user. This recommendation information may indicate the appropriate time for recommending the target item to the user.

[0036] In some embodiments, the recommender may predict the life cycle between the target user and the target item after the current time based on the life cycle of other users of the target item before the current time. For example, the life cycle of other users of the target item before the current time shows the change in other users' level of interest in the target item over time from "awareness" to "interest" to "loyalty." If the life cycle stage of the target user before the current time is "awareness," the recommender may determine that the next life cycle stage of the target user for the target item is "interest" and thereafter "loyalty." In some embodiments, other users may be users with the same characteristics as the target user, for example, users of the same age, which may make the prediction of the life cycle stage of the target user more accurate.

[0037] In some embodiments, the lifecycle of other users of the target item may be retrieved by the recommender from other devices. In some embodiments, the lifecycle of other users of the target item may be generated by the recommender. The recommender may retrieve interaction data between other users and the target item to generate this lifecycle. In some embodiments, the recommender may generate user characteristics based on the predicted lifecycle, wherein the lifecycle used to generate the user characteristics may include a lifecycle phase. In some embodiments, the lifecycle used to generate the user characteristics by the recommender may include multiple lifecycle phases. Accordingly, recommendation information corresponding to multiple lifecycle phases may be generated.

[0038] In some embodiments, the interaction data retrieved by the recommendation device may indicate the interaction information between the target user and multiple items, and the life cycle generated on this basis may indicate the change in the target user's degree of interest in various items over time. For example, the life cycle may be as shown in Fig. 4 may be shown. Fig. Figure 4 shows a schematic representation of a scenario 400 for determining the life cycle in some embodiments of the present disclosure. Scenario 400 includes interaction information 410 between the user and the article, as well as the user's life cycle 420 for the article. Life cycle 420 may include multiple life cycle phases of the user for multiple articles.

[0039] For example, interaction information 411 may indicate that the target user's number of visits to information on the topic "refrigerator" in the period from t1 to t2 is within a first predefined range, and lifecycle phase 421 may indicate that the target user's level of interest in the refrigerator in the period from t1 to t2 is "getting to know" Interaction information 412 may indicate that the target user's number of visits to information on the topic "extractor hood" in the period from t2 to t3 is within a second predefined range, and lifecycle phase 422 may indicate that the target user's level of interest in the extractor hood in the period from t2 to t3 is "interested."Interaction information 413 may indicate that the target user purchased a refrigerator during the period from t3 to t4, and life cycle phase 423 may indicate that the target user's interest level in the refrigerator during the period from t3 to t4 is "purchase." Interaction information 414 may indicate that the target user purchased a range hood during the period from t4 to t5, and life cycle phase 424 may indicate that the target user's interest level in the range hood during the period from t4 to t5 is "purchase."

[0040] In some embodiments, the recommender may generate the target user's life cycle for another item (referred to as the second item for simplicity) based on the target user's life cycle for one item (referred to as the first item for simplicity). The recommender may predict the future life cycle between the target user and the second item based on the life cycle between the target user and the first item prior to the current time. That is, the recommender may predict the user's future level of interest in the second item based on the user's previous interactions with the first item.

[0041] For example, in scenario 400, life cycle 420 may be the target user's life cycle for one or more first items prior to the current time, and the recommender may generate the target user's future life cycle 430 for one or more second items based on this. For example, life cycle 430 may include life cycle phase 431 and life cycle phase 432. Life cycle phase 431 indicates that the target user's degree of interest in a vacuum cleaner is "interest" in the period from t6 to t7 after the current time. Life cycle phase 432 indicates that the target user's degree of interest in a repair tool is "interest" in the period from t8 to t9 after the current time.

[0042] In some embodiments, the recommendation device can also directly generate the life cycle 430 based on the interaction information 410. The recommendation device can generate user characteristics based on the life cycle 430 and generate multiple pieces of recommendation information therefrom. In some embodiments, the recommendation device can generate recommendation information for recommending the item to the user based on the fulfillment of predefined conditions for the user's life cycle phase. For example, the recommendation device can determine that the user's life cycle phase for the vacuum cleaner meets predefined conditions in the period from t6 to t7 and then generate recommendation information for recommending the vacuum cleaner to the user in the period from t6 to t7.The recommendation device can determine that the user's life cycle phase for the repair tool in the period from t8 to t9 meets predefined conditions, and then generate recommendation information for recommending the repair tool in the period from t8 to t9 to the user.

[0043] In some cases, the items already purchased by the user are durable goods, such as refrigerators or other household appliances, for which there is no renewed demand in a short period of time after purchase. In this case, the method provided by the embodiments of the present disclosure can recommend other related items to the user instead of recommending the same item again, thereby meeting the user's needs at the right time and improving the user experience. Furthermore, the solution of the present disclosure can determine a plurality of items to be recommended to the user at different times, generate multiple pieces of recommendation information, and, by executing once, can determine recommendation information for multiple periods, so that the calculation of the recommended items does not need to be performed in real time, thus saving computing resources.

[0044] In some embodiments, the recommender may generate user characteristics for the first item based on the target user's lifecycle and, based on these characteristics, generate recommendation information for recommending the second item to the target user. In some embodiments, the recommender may search for items that have the same characteristics as the first item (e.g., category, price, etc.) and then generate recommendation information for recommending this item to the user.

[0045] In some embodiments, the recommendation device may retrieve historical interaction data of other users regarding the first item and historical interaction data of other users regarding the second item and generate recommendation information based thereon. For example, the user characteristic generated by the recommendation device may indicate that the target user's lifecycle stage for the first item is currently "Purchase." The historical interaction data retrieved by the recommendation device indicates that other users have passed through the "Purchase" lifecycle stage for the second item after the "Purchase" lifecycle stage for the first item, and in this case, the recommendation device may generate recommendation information for recommending the second item to the target user.

[0046] In some embodiments, the user database may contain information about multiple users, and the user-item interaction data may include interaction data between multiple users and multiple items. The recommender may generate multiple life cycles, each associated with one of multiple users, where each life cycle indicates the change in a user's level of interest in an item over time. This allows the recommender to generate multiple pieces of recommendation information, each associated with one of multiple users for item recommendation.

[0047] In some embodiments, the recommender may input the user data and the user's lifecycle for the item into a user encoder to generate embedded users. The recommender may retrieve item data associated with multiple items, input the item data into an item encoder, and generate embedded items. The recommender may input the embedded users and embedded items into a predefined recommendation model and generate recommendation information. For example, the recommendation model may be a two-tower model, where the recommender uses the two-tower model to determine the embedded items most similar to the embedded users to determine the item that can be recommended to the target user, and then generate recommendation information.In some embodiments, the recommender may generate scores for a plurality of items based on the embedded users and embedded items, sort the items based on these scores, and recommend the items to the user according to the sorting.

[0048] In some embodiments, the recommender may evaluate the validity and performance of the recommendation information, for example, by evaluating accuracy, average accuracy, recall rate, precision, normalized discounted cumulative gain (NDCG), or area under the receiver operating characteristic curve (AUC-ROC).

[0049] Fig. 5 shows a block diagram of an article recommendation device 500 according to some embodiments of the present disclosure. As shown in Fig.As shown in Figure 5, the device 500 includes a lifecycle detection module 502 configured to detect a lifecycle, where the lifecycle indicates how the target user's level of interest in an item changes over time. The device 500 further includes a feature generation module 504 configured to generate the target user's user features based on the lifecycle. Furthermore, the device 500 includes a recommendation module 506 configured to generate recommendation information based on the user features to provide the target user with an item recommendation.

[0050] In some embodiments, the lifecycle comprises a plurality of lifecycle phases, wherein the plurality of lifecycle phases indicate the different levels of interest of the target user in an item over a plurality of time periods, and the recommendation module 506 comprises a first recommendation unit configured to generate a plurality of recommendation information based on the user characteristics, wherein the plurality of recommendation information is for recommending a plurality of items to the target user, wherein these plurality of items correspond to the plurality of lifecycle phases.

[0051] In some embodiments, the lifecycle indicates the change in the target user's level of interest in the first item over time, where the first item is an item already purchased by the target user, and the recommendation information is used to recommend a second item to the target user, where the second item is an item not yet purchased by the target user.

[0052] In some embodiments, the recommendation module 506 includes: a historical data acquisition unit configured to retrieve historical data, wherein the historical data includes first historical interaction data of a historical user with the first item and second historical interaction data of a historical user with the second item; and a second recommendation unit configured to generate recommendation information based on the user characteristics, the first historical interaction data, and the second historical interaction data.

[0053] In some embodiments, the first article and the second article have the same attributes.

[0054] In some embodiments, the feature generation module 504 includes: a first lifecycle determination unit configured to determine the target lifecycle based on the lifecycle, wherein the target lifecycle indicates the change in the target user's degree of interest in the second article over time; and a feature generation unit configured to generate user features based on the target lifecycle.

[0055] In some embodiments, the target lifecycle comprises a plurality of lifecycle phases, wherein the plurality of lifecycle phases indicate the different levels of interest of the target user in the second article over a plurality of time periods, and the recommendation module 506 comprises a third recommendation unit configured to generate a plurality of recommendation information based on the user characteristics, wherein the plurality of recommendation information corresponds to the plurality of lifecycle phases.

[0056] In some embodiments, the recommendation module 506 includes: an item feature acquisition unit configured to retrieve the features of a plurality of items; and a fourth recommendation unit configured to generate recommendation information based on the user features, the item features, and a predefined recommendation model.

[0057] In some embodiments, the lifecycle detection module 502 includes: a record retrieval module configured to retrieve a record, the record containing interaction data between the target user and the article; and a second lifecycle determination unit configured to determine the lifecycle based on the record.

[0058] In some embodiments, the lifecycle comprises a plurality of lifecycle phases, wherein the second lifecycle determining unit comprises an interaction count determining unit configured to determine, based on the interaction data, the number of interactions of the target user with the article within a predefined period of time; and a phase determining unit configured to determine, based on the number of interactions, the lifecycle phase of the target user for the article within the predefined period of time.

[0059] In some embodiments, the lifecycle acquisition module 502 includes: a historical cycle retrieval unit configured to retrieve the historical lifecycle, wherein the historical lifecycle indicates the change in a historical user's degree of interest in an article over time; and a third lifecycle determination unit configured to determine the lifecycle based on the historical lifecycle.

[0060] Fig.6 shows a schematic block diagram of an example device 600 that can be used to implement an embodiment of the present disclosure. As shown in the figure, the device 600 includes a computing unit 601 that can perform various suitable actions and processes according to computer program instructions stored in a read-only memory (ROM) 602 or loaded from a storage unit 608 into a random access memory (RAM) 603. The RAM 603 can also store various programs and data required for the operation of the device 600. The computing unit 601, the ROM 602, and the RAM 603 are interconnected via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0061] Several components in the device 600 are connected to the I / O interface 605, including: an input unit 606, such as a keyboard, mouse, etc.; an output unit 607, such as various types of displays, speakers, etc.; a storage unit 608, such as a hard disk, optical storage, etc.; and a communication unit 609, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 609 enables the device 600 to exchange information / data with other devices via computer networks such as the Internet and / or various telecommunications networks.

[0062] Computing unit 601 may be any type of general-purpose and / or application-specific processing unit with processing and computing capabilities. Examples of computing unit 601 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various application-specific computing chips for artificial intelligence (AI), various computing units executing algorithms for machine learning models, digital signal processors (DSPs), and any suitable processors, controllers, and microcontrollers. Computing unit 601 performs various of the methods and processes described above, such as method 200. In some embodiments, method 200 may be implemented, for example, as a computer software program physically embodied in a machine-readable medium such as storage unit 608.In some embodiments, the computer program may be loaded and / or installed in whole or in part onto the device 600 via the ROM 602 and / or the communication unit 609. When the computer program is loaded into the RAM 603 and executed by the computing unit 601, one or more steps of the method 200 described above may be performed. Alternatively, in other embodiments, the computing unit 601 may be configured to execute the method 200 in any other suitable manner (e.g., via firmware).

[0063] The functions described above may be performed, at least in part, by one or more hardware logic components. Examples of hardware logic components that may be used include, for example, and without limitation, field programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-chips (SOCs), programmable logic devices (CPLDs), etc.

[0064] The program code for implementing the method of the present disclosure may be written using one or any combination of several programming languages. This program code may be provided to a processor or controller of a general-purpose computer, an application-specific computer, or other programmable data processing device such that, when the program code is executed by the processor or controller, the functions / operations specified in the flowcharts and / or block diagrams are performed. The program code may be executed entirely on one machine, partially on one machine, as a standalone software package, partially on one machine and partially on a remote machine, or entirely on a remote machine or server.

[0065] In the context of the present disclosure, the machine-readable medium may be a physical medium embodying or storing programs for use by, or in connection with, a system, apparatus, or device to execute instructions. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media may include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or apparatus, or any suitable combination thereof.More specific examples of machine-readable storage media include electrical connections based on one or more wires, a portable computer drive, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. Although the operations are presented in a particular order, this does not imply that these operations must be performed in the order shown or sequentially, or that all of the operations presented must be performed to achieve the desired results. In certain environments, multitasking and parallel processing may be advantageous.Although several specific implementation details are included in the above discussion, these should not be construed as limiting the scope of the present disclosure. Certain features described in the context of separate embodiments may also be implemented in combination in a single implementation. Conversely, various functions described in the context of a single implementation may also be implemented in multiple implementations separately or in any suitable subcombination.

[0066] Although the subject matter has been described using language specifically intended for the structural features and / or logical acts of the method, it should be understood that the subject matter defined by the appended claims is not limited to the specific features or acts described above. On the contrary, the specific features and acts described above merely embody exemplary embodiments of the claims.

Claims

[1] Article recommendation procedure, comprising: Capturing a lifecycle, where the lifecycle indicates how a target user's level of interest in an item changes over time; Generating user characteristics of the target user based on the life cycle; and Generating recommendation information based on user characteristics to provide product recommendations to the target user. [2] The method of claim 1, wherein the lifecycle comprises a plurality of lifecycle phases, and the plurality of lifecycle phases indicate the plurality of levels of interest of the target user in the item during the plurality of time periods, and wherein generating recommendation information based on the user characteristics to provide the target user with an item recommendation comprises: Generating multiple recommendation information based on the user characteristics, wherein the multiple recommendation information is used to recommend multiple items to the target user, and the multiple items correspond to the multiple life cycle stages. [3] The method of claim 1, wherein the life cycle indicates how the target user's level of interest in a first item changes over time, wherein the recommendation information is used to recommend a second item to the target user, and wherein the second item is different from the first item. [4] The method of claim 3, wherein generating recommendation information based on the user characteristics to provide an item recommendation to the target user comprises: Collecting historical data, wherein the historical data comprises first historical interaction data of a historical user with the first item and second historical interaction data of the historical user with the second item; and Generate recommendation information based on user characteristics, first historical interaction data, and second historical interaction data. [5] The method of claim 3, wherein the first article and the second article have the same attributes. [6] The method of claim 3, wherein generating the user characteristics of the target user based on the life cycle comprises: Determining a target lifecycle based on the lifecycle, where the target lifecycle indicates how the target user's level of interest in the second article changes over time; and Generating user characteristics based on the target lifecycle. [7] The method of claim 6, wherein the target lifecycle comprises a plurality of lifecycle phases, and the plurality of lifecycle phases indicate a plurality of levels of interest of the target user in the second item during the plurality of time periods, and wherein generating recommendation information based on the user characteristics to provide the target user with an item recommendation comprises: Generating multiple recommendation information based on the user characteristics, wherein the multiple recommendation information each corresponds to the multiple life cycle stages. [8] The method of claim 1, wherein generating recommendation information based on the user characteristics to provide an item recommendation to the target user comprises: Recording item characteristics of multiple items; and Generation of recommendation information based on user characteristics, item characteristics and a predefined recommendation model. [9] A method according to any one of claims 1 to 8, wherein the life cycle detection comprises: Collecting a data set, the data set including interaction data between the target user and the article; and Determine the life cycle based on the dataset. [10] The method of claim 9, wherein the life cycle comprises a plurality of life cycle phases and determining the life cycle based on the data set comprises: Determining the number of interactions between the target user and the article within a predefined period based on the interaction data; and Determine the target user's lifecycle stage for the item within the predefined period based on the number of interactions. [11] A method according to any one of claims 1 to 8, wherein the life cycle detection comprises: Capturing a historical lifecycle, where the historical lifecycle indicates how a historical user's level of interest in the item changes over time; and Determination of the life cycle based on the historical life cycle. [12] Device for article recommendation, comprising: a lifecycle capture module configured to capture a lifecycle, wherein the lifecycle indicates how the target user's level of interest in an item changes over time; a feature generation module configured to generate the user features of the target user based on the lifecycle; and a recommendation module configured to generate recommendation information based on the user characteristics to provide an item recommendation to the target user. [13] Electronic device comprising: at least one processor; and a memory coupled to the at least one processor and having instructions stored thereon, the instructions, when executed by the at least one processor, causing the device to perform a method according to any one of claims 1 to 11. [14] A computer program product physically stored on a non-transitory, computer-readable medium and comprising machine-executable instructions, the machine-executable instructions, when executed, causing the machine to perform the steps of the method of any one of claims 1 to 11.

Images