A commodity personalization recommendation method and device, a storage medium, and an electronic device

Abstract

The application provides a personalized commodity recommendation method and device, a storage medium and electronic equipment, and is applied to the technical field of Internet application.The application responds to a user commodity recommendation trigger operation, asynchronously loads preloaded commodity recommendation data in a local cache, uploads user behavior data and user preference parameters to a cloud recommendation engine, and obtains customized commodity recommendation data.Through comparison of the customized commodity recommendation data and the preloaded commodity recommendation data by using a unique identifier and additional information of the commodity, the application identifies commodities to be added, updated or deleted, synchronizes a personalized commodity recommendation list and the local cache, and displays the commodities to the user.Through real-time comparison and fusion of the cloud customized commodity recommendation data and the local preloaded commodity recommendation data, the application dynamically updates the personalized commodity recommendation list, improves the loading speed and matching accuracy of the commodity recommendation data, optimizes the commodity recommendation process, and enhances the personalized effect and overall user experience.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of Internet application, and in particular to a personalized commodity recommendation method and device, a storage medium and an electronic device. BACKGROUND

[0002] With the rapid development of e-commerce, the user's requirements for page loading speed and data real-time performance are increasing in the process of commodity page display of e-commerce APP (Application). However, the current e-commerce APP has problems such as slow first loading speed, untimely data updating, and repeated or missing page data in the commodity display link. These problems not only affect the user's browsing experience, but also may lead to a decline in recommendation effect, thereby affecting the platform's conversion rate and user stickiness.

[0003] Traditional solutions mostly rely on static content delivery network (CDN) caching or single data request mode. Although this method can reduce the server pressure to some extent, due to the static nature of caching, it is difficult to meet the real-time updating requirements of dynamic user demand and personalized recommendation. Especially in the case of network fluctuations or high concurrency access, the traditional method cannot effectively guarantee the accuracy and consistency of data, which may lead to page information lag or incompleteness, thereby affecting user experience and system stability.

[0004] Therefore, how to optimize the data loading and updating of commodity recommendation and improve the personalized effect of commodity recommendation and the overall user experience has become a technical problem that needs to be solved by technical personnel in the field. SUMMARY

[0005] In view of the above problems, the present application provides a personalized commodity recommendation method, device, storage medium and electronic device to overcome the above problems or at least partially solve the above problems, and the technical solution is as follows:

[0006] A personalized commodity recommendation method, comprising:

[0007] Asynchronously loading preloaded commodity recommendation data in a local cache in a personalized commodity recommendation list in response to a user's commodity recommendation triggering operation, wherein the preloaded commodity recommendation data includes at least one unique commodity identifier and additional information of a commodity;

[0008] Uploading user behavior data and user preference parameters of the user to a cloud recommendation engine to obtain customized commodity recommendation data returned by the cloud recommendation engine, wherein the customized commodity recommendation data includes at least one unique commodity identifier and additional information of a commodity;

[0009] The customized product recommendation data is compared with the preloaded product recommendation data by using the unique product identifier and additional information to identify to-be-fused products, wherein the to-be-fused products include to-be-added products, to-be-updated products and / or to-be-deleted products;

[0010] Based on the to-be-fused products, data synchronization is performed on the personalized product recommendation list and the local cache, and recommended products are displayed to the user according to the synchronized personalized product recommendation list.

[0011] Optionally, the data synchronization performed on the personalized product recommendation list and the local cache based on the to-be-fused products comprises:

[0012] For any to-be-added product, the to-be-added product is inserted into the personalized product recommendation list at a position matched with a recommendation score of the to-be-added product according to the recommendation score of the to-be-added product, and a unique product identifier and additional information of the to-be-added product are saved to the local cache;

[0013] For any to-be-updated product, real-time product key indicator data of the to-be-updated product in the customized product recommendation data is compared with historical product key indicator data of the to-be-updated product in the preloaded product recommendation data, and if the real-time product key indicator data is better than the historical product key indicator data, the real-time product key indicator data replaces the historical product key indicator data of the to-be-updated product in the personalized product recommendation list and the local cache;

[0014] For any to-be-deleted product, whether an activity level of the to-be-deleted product is lower than a preset activity level threshold is detected, and if yes, the to-be-deleted product is deleted from the personalized product recommendation list and the local cache if the to-be-deleted product meets a preset deletion condition.

[0015] Optionally, in the process of the data synchronization performed on the personalized product recommendation list and the local cache based on the to-be-fused products, the method further comprises:

[0016] Each product is divided into a plurality of product segment intervals according to a unique product identifier, and a lock object is initialized for each product segment interval;

[0017] A plurality of threads are started, and each thread is specified to process data synchronization of the to-be-fused products in one or more product segment intervals;

[0018] In a case that the lock object of the commodity segment interval is currently not occupied, the thread acquires the lock object, starts to traverse the data synchronization of the to-be-fused commodities in the commodity segment interval, and releases the lock object after the data synchronization of the to-be-fused commodities in the commodity segment interval.

[0019] In a case that the lock object of the commodity segment interval is currently occupied, the thread waits for the lock object to be released.

[0020] Optionally, in the process of synchronizing the personalized commodity recommendation list and the local cache based on the to-be-fused commodities, the method further comprises:

[0021] Based on the additional information of the to-be-fused commodities, it is detected whether there is data anomaly in the to-be-fused commodities, and in a case that it is detected that any to-be-fused commodity has data anomaly, a rollback operation is performed on the personalized commodity recommendation list and the local cache, so as to restore the personalized commodity recommendation list and the local cache to a state before the current data synchronization.

[0022] Optionally, in the process of synchronizing the personalized commodity recommendation list and the local cache based on the to-be-fused commodities, the method further comprises:

[0023] In real time, it is monitored whether network interruption occurs in the data synchronization process, and if so, a rollback operation is performed on the personalized commodity recommendation list and the local cache, so as to restore the personalized commodity recommendation list and the local cache to a state before the current data synchronization.

[0024] Optionally, the rollback operation comprises:

[0025] In the order of the recommendation scores of the commodities in the preloaded commodity recommendation data from high to low before the current data synchronization, data in the personalized commodity recommendation list and the local cache is restored in sequence, so that the preloaded commodity recommendation data in the restored personalized commodity recommendation list and the local cache is consistent with that before the current data synchronization.

[0026] Optionally, the method further comprises:

[0027] It is detected whether data of each commodity in the local cache is valid, and in a case that any commodity has invalid data, the step of uploading user behavior data and user preference parameters of the user to a cloud recommendation engine and obtaining customized commodity recommendation data returned by the cloud recommendation engine is performed.

[0028] The application discloses a personalized commodity recommendation device, which comprises a local data loading unit, a cloud customized data obtaining unit, a to-be-fused commodity identifying unit and a data synchronization unit.

[0029] The local data loading unit is configured to load preloaded commodity recommendation data in a local cache in a personalized commodity recommendation list in response to a commodity recommendation triggering operation of a user, wherein the preloaded commodity recommendation data comprises unique commodity identifiers and additional information of at least one commodity.

[0030] The cloud customized data obtaining unit is configured to upload user behavior data and user preference parameters of the user to a cloud recommendation engine and obtain customized commodity recommendation data returned by the cloud recommendation engine, wherein the customized commodity recommendation data comprises unique commodity identifiers and additional information of at least one commodity.

[0031] The to-be-fused commodity identifying unit is configured to compare the customized commodity recommendation data with the preloaded commodity recommendation data by using the unique commodity identifiers and the additional information, and identify to-be-fused commodities, wherein the to-be-fused commodities comprise to-be-added commodities, to-be-updated commodities and / or to-be-deleted commodities.

[0032] The data synchronization unit is configured to perform data synchronization on the personalized commodity recommendation list and the local cache based on the to-be-fused commodities, and display recommendation commodities to the user according to the synchronized personalized commodity recommendation list.

[0033] A computer readable storage medium, which stores a program, wherein the program is executed by a processor to implement the personalized commodity recommendation method.

[0034] An electronic device, comprising at least one processor and at least one memory connected with the processor, wherein the processor and the memory complete mutual communication through a bus; the processor is configured to call program instructions in the memory to execute the personalized commodity recommendation method.

[0035] By means of the technical scheme, the application provides a commodity personalized recommendation method, device, storage medium and electronic equipment, in response to a commodity recommendation triggering operation of a user, preloaded commodity recommendation data in a local cache is loaded asynchronously in a personalized commodity recommendation list, wherein the preloaded commodity recommendation data comprises unique commodity identifiers and additional information of at least one commodity; user behavior data and user preference parameters of the user are uploaded to a cloud recommendation engine to obtain customized commodity recommendation data returned by the cloud recommendation engine, wherein the customized commodity recommendation data comprises unique commodity identifiers and additional information of at least one commodity; the customized commodity recommendation data is compared with the preloaded commodity recommendation data by means of the unique commodity identifiers and the additional information to identify to-be-fused commodities, wherein the to-be-fused commodities comprise to-be-added commodities, to-be-updated commodities and / or to-be-deleted commodities; based on the to-be-fused commodities, data synchronization is performed on the personalized commodity recommendation list and the local cache, and recommended commodities are displayed to the user according to the synchronized personalized commodity recommendation list. The application realizes real-time comparison and fusion of locally preloaded data and cloud customized recommendation data by means of asynchronous loading, dynamically updates the personalized commodity recommendation list, realizes quick response and accurate matching of commodity recommendation data, and thus optimizes the data loading and updating process of commodity recommendation and improves the personalized effect of commodity recommendation and the overall user experience.

[0036] The above description is only a summary of the technical scheme of the application, in order to more clearly understand the technical means of the application, the application can be implemented according to the content of the specification, and in order to make the above and other purposes, characteristics and advantages of the application more obvious and easy to understand, the following specific embodiments of the application are described. BRIEF DESCRIPTION OF DRAWINGS

[0037] Various other advantages and benefits will become apparent to those of ordinary skill in the art, upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included to provide a description of preferred embodiments, and are not meant to limit the present application. Furthermore, the same reference numerals are used throughout the several drawings to refer to same or like parts. In the drawings:

[0038] Figure 1 A flowchart of an embodiment of a commodity personalized recommendation method provided by the application is shown;

[0039] Figure 2 A structure diagram of a commodity personalized recommendation device provided by the application is shown;

[0040] Figure 3 A structure diagram of an electronic equipment provided by the application is shown. DETAILED DESCRIPTION

[0041] Exemplary embodiments of the present application will be described herein below with reference to the accompanying drawings. While exemplary embodiments of the present application are shown in the drawings, it is understood that the present application can be embodied in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the present application to those skilled in the art.

[0042] As shown in FIG. 1, an embodiment of the present application provides a flowchart of an embodiment of a method for personalized product recommendation, which can include the following steps: Figure 1

[0043] S100, in response to a product recommendation trigger operation of a user, asynchronously loading preloaded product recommendation data in a local cache in a personalized product recommendation list, wherein the preloaded product recommendation data includes at least one unique product identifier and additional information of a product.

[0044] Wherein, the product recommendation trigger operation refers to an operation or event performed by the user in the APP (Application) to trigger personalized product recommendation, in order to prepare to display or update the personalized product recommendation list. For example, the product recommendation trigger operation can include the user entering the product detail page, starting the APP, browsing a certain product category page, etc.

[0045] Wherein, the personalized product recommendation list refers to a product collection list specially generated by a recommendation algorithm according to the user's interests, behaviors, preferences, and real-time access scenarios. The products in the personalized product recommendation list can be dynamically sorted according to the user portrait and the context environment. The present embodiment ensures the response speed and display effect of the personalized product recommendation list through asynchronous loading and cache management.

[0046] Wherein, the preloaded product recommendation data refers to product recommendation related information prepared in advance from the local cache or the background asynchronously before the actual demand of the user. For example, through iOS background task (such as Background Fetch or NSURLSession background transmission) asynchronous loading. The preloaded product recommendation data contains unique product identifiers and additional information, which aims to speed up the display speed and improve the fluency of the recommended products. The preloaded product recommendation data usually also includes product basic information (such as name, price) and high-definition product pictures, etc.

[0047] Wherein, the local cache can be NSCache or disk cache which supports fast reading and display.

[0048] Wherein, the unique product identifier refers to a marker used to uniquely identify a certain product, which can be a string or a number code, such as SKU (stock keeping unit number) and product ID, etc.

[0049] ​The additional information refers to product metadata stored corresponding to the unique product identifier, and is mainly used to describe the state, version and validity of the product recommendation data. Common fields include timestamp and version number, etc.

[0050] S110, upload the user behavior data and user preference parameters of the user to the cloud recommendation engine to obtain customized product recommendation data returned by the cloud recommendation engine, wherein the customized product recommendation data includes at least one unique product identifier of a product and additional information.

[0051] The user behavior data refers to various interaction information and operation records generated by the user on the APP or platform, reflecting the user's interest, preference and use habit. The user behavior data can include the user's historical browsing records, click behavior, search keywords, purchase records, stay time, collection and likes related to the product.

[0052] The user preference parameter refers to various attributes and indicators used to describe and quantify the user's interest, habit, hobby and specific needs. The user preference parameter can include the interest label (such as favorite product category and brand) provided by the user, the summary of historical behavior (such as frequently visited product type and price range) and user demographic information (such as age, gender and region), etc. It can also include potential interest points obtained by analyzing user behavior through machine learning models.

[0053] The cloud recommendation engine refers to an intelligent recommendation system deployed on a cloud server, which is used to receive user behavior data and user preference parameters, combine real-time hot information, analyze the products that the user has purchase intention and rapid sales growth in the near future through big data analysis and machine learning algorithms, and generate customized product recommendation data in real time.

[0054] The cloud recommendation model can perform preprocessing including data cleaning and formatting after receiving the user behavior data and user preference parameters, analyze the user activity, category preference, price sensitivity and other features from the user behavior data and user preference parameters, and then construct a multi-dimensional user-product interaction matrix by fusing real-time hot information to construct a hot weight index. Then, through similarity calculation, the product with high user feature matching degree is recommended, and the recommendation score of the product is output using a deep learning recommendation network model, and finally the customized product recommendation data containing the unique product identifier and additional information of the product is formed.

[0055] The customized product recommendation data refers to the personalized product recommendation result set returned by the cloud recommendation engine to the client according to the user behavior data and preference parameters.

[0056] S120, comparing the customized product recommendation data with the preloaded product recommendation data by using the unique product identifier and the additional information to identify the to-be-fused products, wherein the to-be-fused products include to-be-added products, to-be-updated products and / or to-be-deleted products.

[0057] The to-be-fused products refer to a set of products that need to be subjected to data synchronization operations after the customized product recommendation data is compared with the preloaded product recommendation data on the client side.

[0058] The to-be-added products refer to products that exist in the customized product recommendation data but do not exist in the preloaded product recommendation data and need to be newly displayed in the personalized product recommendation list.

[0059] The to-be-updated products refer to products that exist in both the customized product recommendation data and the preloaded product recommendation data, but the display recommendation data changes due to the additional information (such as version number and timestamp), and the corresponding product in the preloaded product recommendation data needs to be replaced.

[0060] The to-be-deleted products refer to products that exist in the preloaded product recommendation data but do not exist in the customized product recommendation data and need to be removed from the preloaded product recommendation data.

[0061] Specifically, the embodiment of the application can take the preloaded product recommendation data as a basic data set, traverse the customized product recommendation data, and compare the unique product identifier piece by piece to identify the to-be-fused products based on the unique product identifier and the additional information.

[0062] S130, based on the to-be-fused products, performing data synchronization on the personalized product recommendation list and the local cache, and displaying the recommended products to the user according to the synchronized personalized product recommendation list.

[0063] Specifically, the embodiment of the application can start a concurrent thread, for the to-be-added products marked as added, directly append them to the personalized product recommendation list and the local cache in the thread to enrich the product recommendation content of the personalized product recommendation list, for the to-be-updated products matched and marked as updated, replace the corresponding product information in the personalized product recommendation list and the local cache in the thread to ensure that the data is the latest, and for the products marked as deleted, remove the corresponding product item from the personalized product recommendation list and the local cache in the thread to ensure the accuracy and effectiveness of the data.

[0064] Specifically, after the personalized commodity recommendation list is synchronized with the locally cached data, the embodiment of the present application can use the synchronized personalized commodity recommendation list as the latest data source to drive the interface rendering module to quickly update the UI (User Interface) page used for commodity display. Meanwhile, the offline caching mechanism is adopted to ensure that the user can still obtain stable and continuous recommended commodity browsing experience in the case of network exception or disconnection.

[0065] The present application provides a personalized commodity recommendation method, which comprises: in response to a commodity recommendation trigger operation of a user, asynchronously loading preloaded commodity recommendation data in a local cache in a personalized commodity recommendation list, wherein the preloaded commodity recommendation data comprises unique commodity identifiers and additional information of at least one commodity; uploading user behavior data and user preference parameters of the user to a cloud recommendation engine to obtain customized commodity recommendation data returned by the cloud recommendation engine, wherein the customized commodity recommendation data comprises unique commodity identifiers and additional information of at least one commodity; using the unique commodity identifiers and the additional information, comparing the customized commodity recommendation data with the preloaded commodity recommendation data to identify to-be-fused commodities, wherein the to-be-fused commodities comprise to-be-added commodities, to-be-updated commodities and / or to-be-deleted commodities; based on the to-be-fused commodities, performing data synchronization on the personalized commodity recommendation list and the local cache, and showing recommended commodities to the user according to the synchronized personalized commodity recommendation list. The present application realizes dynamic updating of the personalized commodity recommendation list through real-time comparison and fusion of the locally preloaded data and the cloud customized recommendation data, realizes quick response and accurate matching of commodity recommendation data, thereby optimizing the data loading and updating process of commodity recommendation and improving the personalized effect of commodity recommendation and the overall user experience.

[0066] Optionally, for any to-be-added commodity, the embodiment of the present application inserts the to-be-added commodity into a position matched with a recommendation score in the personalized commodity recommendation list according to the recommendation score of the to-be-added commodity, and saves the unique commodity identifier and the additional information of the to-be-added commodity to the local cache.

[0067] Specifically, for any to-be-added commodity, the embodiment of the present application can insert the commodity into a suitable position matched with the score of the commodity in the personalized commodity recommendation list according to the recommendation score of the commodity, thereby ensuring that high-score commodities are preferentially displayed and improving the recommendation effect. Meanwhile, the unique commodity identifier and the additional information of the new commodity are saved to the local cache, thereby guaranteeing the consistency and persistence of data and supporting subsequent quick access and offline use.

[0068] Optionally, for any to-be-updated commodity, the application compares whether the real-time commodity key indicator data of the to-be-updated commodity in the customized commodity recommendation data is better than the historical commodity key indicator data in the preloaded commodity recommendation data, and if so, replaces the historical commodity key indicator data of the to-be-updated commodity in the personalized commodity recommendation list and the local cache with the real-time commodity key indicator data.

[0069] The commodity key indicator data refers to core data information for measuring and describing the state, performance and user feedback of a commodity. The commodity key indicator data can include at least one of the price, inventory, user evaluation, sales rate, click rate, conversion rate and return rate of the commodity.

[0070] Specifically, for any to-be-updated commodity, the application can compare the real-time commodity key indicator data of the commodity in the customized commodity recommendation data with the historical commodity key indicator data in the preloaded commodity recommendation data. If the real-time commodity key indicator data is better than the historical commodity key indicator data, the real-time commodity key indicator data is replaced with the historical commodity key indicator data of the corresponding commodity in the personalized commodity recommendation list and the local cache, ensuring that the recommended content reflects the latest and better commodity state, and synchronously updating the related information in the local cache to maintain the consistency and timeliness of the data.

[0071] Optionally, for any to-be-deleted commodity, the application detects whether the activity level of the to-be-deleted commodity is lower than a preset activity level threshold, and if so, deletes the to-be-deleted commodity in the personalized commodity recommendation list and the local cache if the to-be-deleted commodity meets a preset deletion condition.

[0072] The activity level refers to the frequency and intensity of user interaction with a commodity within a certain time range, and is used to reflect the degree of attention and demand of the commodity by users. The application can obtain the activity level by calculating indicators such as the number of clicks, the number of views, the number of purchases, the number of collections, etc.

[0073] Specifically, for any to-be-deleted commodity, the application can detect whether the activity level of the commodity in recent user interaction is lower than a preset activity level threshold. If the activity level is lower than the threshold and meets a preset deletion condition (such as long-term no clicks, extremely low sales or zero inventory, etc.), the commodity is deleted from the personalized commodity recommendation list and the local cache, releasing storage resources and optimizing the recommendation efficiency and response speed.

[0074] Optionally, in the above Figure 1 Based on one or more corresponding embodiments, in another optional embodiment of the application, in the process of synchronizing the data of the personalized commodity recommendation list and the local cache based on the to-be-merged commodity, the method can further include:

[0075] According to the unique commodity identifier, each commodity is divided into a plurality of commodity segment intervals, and a lock object is initialized for each commodity segment interval; a plurality of threads are started, each thread is designated to process data synchronization of one or more commodities to be merged in a commodity segment interval; when a thread is ready to start processing data synchronization of commodities to be merged in a commodity segment interval, if the lock object of the commodity segment interval is currently unoccupied, the lock object is obtained, the processing of data synchronization of commodities to be merged in the commodity segment interval is started, and after the data synchronization of commodities to be merged in the commodity segment interval, the lock object is released; if the lock object of the commodity segment interval is currently occupied, the thread waits for the lock object to be released.

[0076] Among them, the commodity segment interval refers to dividing the entire commodity set into several segments according to the unique commodity identifier (such as commodity ID), and each interval contains commodities within a certain range.

[0077] Among them, the lock object is a kind of concurrent control mechanism, which is used to ensure mutual exclusion access to data in the commodity segment interval in the multi-thread environment. Each commodity segment interval corresponds to an independent lock object. When a thread needs to operate the data in the interval, the lock object must be obtained first to prevent other threads from modifying at the same time, and to ensure the atomicity and consistency of data operation.

[0078] For ease of understanding, by way of example, it is assumed that there are 1000 to-be-merged commodities, the commodity IDs are 1 to 1000, and the to-be-merged commodities need to be processed synchronously in parallel. The commodity IDs are divided into 5 commodity segmented intervals according to intervals: interval A (commodity ID 1~200), interval B (commodity ID 201~400), interval C (commodity ID 401~600), interval D (commodity ID 601~800), and interval E (commodity ID 801~1000). An independent lock object is created for each interval: LockA, LockB, LockC, LockD, and LockE. Three threads are started: thread 1, thread 2, and thread 3. Thread 1 is responsible for processing commodity data in interval A and interval B, thread 2 is responsible for processing commodity data in interval C and interval D, and thread 3 is responsible for processing commodity data in interval E. When thread 1 is ready to process data in interval A, it checks whether LockA is occupied. If LockA is idle, thread 1 acquires LockA and starts traversing all to-be-merged commodities in interval A to perform data synchronization. After processing is completed, thread 1 releases LockA. If LockA is occupied by another thread at this time (for example, another thread also attempts to process interval A), thread 1 will wait until LockA is released before acquiring the lock. Similarly, thread 1 processes data in interval B by first acquiring LockB and then releasing it after completion. Thread 2 processes interval C and interval D in turn by acquiring LockC and LockD, respectively, to ensure that the same interval is not operated by multiple threads at the same time. Thread 3 processes interval E by acquiring LockE for synchronization and releasing it after completion.

[0079] Embodiments of the present application divide to-be-merged commodities into multiple segmented intervals according to unique commodity identifiers, initialize independent lock objects for each interval, and start multiple threads to process to-be-merged commodities in different segmented intervals, thereby achieving fine-grained synchronization control. When a thread processes data in a certain segmented interval, it must first acquire the lock of the interval to ensure that only one thread accesses the interval at the same time, effectively avoiding data competition and inconsistency problems. If the lock is occupied, the thread waits for the lock to be released to ensure the correctness and atomicity of the synchronization operation. This not only improves the efficiency of concurrent processing and reduces synchronization delay, but also ensures the consistency and integrity of the personalized commodity recommendation list and the local cache, significantly improving the response speed and stability of the commodity recommendation service, and ultimately presenting more accurate and timely recommendation content to users.

[0080] Optionally, in the above Figure 1 Based on one or more embodiments, another optional embodiment provided by embodiments of the present application includes the following steps in the process of synchronizing data between the personalized commodity recommendation list and the local cache based on to-be-merged commodities:

[0081] Based on the additional information of the to-be-fused commodity, it is detected whether the to-be-fused commodity has data anomaly, and in a case that any to-be-fused commodity is detected to have data anomaly, a rollback operation is performed on the personalized commodity recommendation list and the local cache, so that the personalized commodity recommendation list and the local cache are restored to a state before the current data synchronization.

[0082] Specifically, the embodiment of the present application can determine that the to-be-fused commodity has data anomaly when the time stamp of the to-be-fused commodity exceeds a preset reasonable range of time, the version number is discontinuous, or the data format does not meet the expectation. The to-be-fused commodity can also be determined to have data anomaly in combination with the update frequency and the user access frequency of the to-be-fused commodity. For example, for a high-frequency update commodity, if the time stamp exceeds the set threshold and the version number lags behind, the data is considered to be expired and determined to be abnormal. For a low-frequency access commodity, if data loss or format error is detected, the data is determined to be abnormal. In a case that any to-be-fused commodity is detected to have data anomaly, a rollback operation is triggered immediately to restore the personalized commodity recommendation list and the local cache to a stable state before the current data synchronization.

[0083] When any data anomaly is detected based on the additional information (such as time stamp, version number, etc.) of the to-be-fused commodity, the embodiment of the present application can trigger a rollback operation in time to restore the personalized commodity recommendation list and the local cache to a stable state before the current data synchronization, thereby effectively preventing abnormal data from causing errors in the personalized commodity recommendation list and the local cache, ensuring the integrity and stability of the commodity recommendation data, improving the continuity and accuracy of the user experience, avoiding the risk of affecting the recommendation effect due to data anomaly, and significantly enhancing the reliability of the entire commodity recommendation service and user satisfaction.

[0084] Optionally, in the above Figure 1 Based on one or more embodiments, in another optional embodiment provided by the embodiment of the present application, in the process of synchronizing data of the personalized commodity recommendation list and the local cache based on the to-be-fused commodity, the method can further include:

[0085] The data synchronization process is monitored in real time to determine whether network interruption occurs, and if so, a rollback operation is performed on the personalized commodity recommendation list and the local cache, so that the personalized commodity recommendation list and the local cache are restored to a state before the current data synchronization.

[0086] Specifically, in the data synchronization process, the embodiment of the present application will monitor the network connection state in real time to timely find abnormal conditions such as network interruption. Once the network interruption is detected, it is immediately determined that the current data synchronization update fails, and it is considered that the current synchronized data has the risk of incompleteness or loss. In order to protect the data integrity and accuracy of the personalized commodity recommendation list and the local cache, a rollback operation is immediately triggered to restore the personalized commodity recommendation list and the local cache to a stable state before the start of the current data synchronization, thereby avoiding the half update or error data caused by network anomalies from being displayed to the user, thereby effectively protecting the coherence of the recommended content and the stability of the user experience, and ensuring that the commodity recommendation service can still provide reliable service under network fluctuations.

[0087] Optionally, the rollback operation provided by the embodiment of the present application can include: restoring the data in the personalized commodity recommendation list and the local cache in the order of the recommendation scores of the preloaded commodity recommendation data from high to low before the current data synchronization, so that the preloaded commodity recommendation data in the restored personalized commodity recommendation list and the local cache is consistent with that before the current data synchronization.

[0088] In the rollback operation process, in order to ensure that the personalized commodity recommendation list and the local cache can be accurately and orderly restored to the state before the current data synchronization, the data of the corresponding commodity is restored in the order of the recommendation scores of the preloaded commodity recommendation data from high to low, thereby ensuring that the popular commodities and the commodities recently concerned by the user are preferentially restored, effectively reducing the user experience fluctuation caused by data rollback, making the preloaded commodity recommendation data in the restored personalized commodity recommendation list and the local cache consistent with that before synchronization, maintaining the continuity and stability of the recommended content, and maximizing the protection of the user to obtain high-quality personalized recommendation service.

[0089] Optionally, in the above Figure 1 Based on one or more embodiments corresponding thereto, another optional embodiment provided by the embodiment of the present application can further include:

[0090] Detecting the data validity of each commodity in the local cache, and in the case that any commodity has invalid data, uploading the user behavior data and user preference parameters of the user to the cloud recommendation engine to obtain the customized commodity recommendation data returned by the cloud recommendation engine.

[0091] The embodiment of the present application can periodically detect the validity of the data of each commodity in the local cache during the continuous operation of the commodity recommendation service, to verify whether the data is complete, accurate and expired. Once any commodity data is found to be invalid, abnormal or expired, the subsequent processing flow is triggered immediately: uploading the user behavior data and user preference parameters of the user to the cloud recommendation engine, obtaining the customized commodity recommendation data returned by the cloud recommendation engine, so as to ensure the personalization and real-time of the commodity recommendation content. The embodiment of the present application closely cooperates with the cache reconstruction task and the data fusion algorithm, and can quickly correct the inconsistency or obsolete information of the local cache in the network fluctuation or high concurrency access scene, so as to ensure that the user always gets the latest and most accurate recommendation display, and improve the overall user experience and stability of the recommendation service.

[0092] Although the operations are depicted in a particular order, this should not be understood as requiring the operations to be performed in the particular order shown or in sequential order. Under certain circumstances, multitasking and parallel processing can be advantageous.

[0093] It should be understood that each of the steps described in the method embodiments of the present application can be performed in different order and / or in parallel. In addition, the method embodiments can include additional steps and / or omit the execution of the steps shown. The scope of the present application is not limited in this respect.

[0094] Corresponding to the above-mentioned method embodiment, the embodiment of the present application also provides a commodity personalized recommendation device, as shown in Figure 2 The structure can include: a local data loading unit 10, a cloud customized data obtaining unit 20, a to-be-fused commodity identification unit 30 and a data synchronization unit 40.

[0095] The local data loading unit 10 is used for asynchronously loading the preloaded commodity recommendation data in the local cache in the personalized commodity recommendation list in response to the commodity recommendation triggering operation of the user, wherein the preloaded commodity recommendation data includes the unique commodity identifier and additional information of at least one commodity.

[0096] The cloud customized data obtaining unit 20 is used for uploading the user behavior data and user preference parameters of the user to the cloud recommendation engine, and obtaining the customized commodity recommendation data returned by the cloud recommendation engine, wherein the customized commodity recommendation data includes the unique commodity identifier and additional information of at least one commodity.

[0097] The to-be-fused commodity identification unit 30 is used for comparing the customized commodity recommendation data with the preloaded commodity recommendation data by using the unique commodity identifier and the additional information, and identifying the to-be-fused commodity, wherein the to-be-fused commodity includes the to-be-added commodity, the to-be-updated commodity and / or the to-be-deleted commodity.

[0098] The data synchronization unit 40 is configured to synchronize the personalized product recommendation list and the local cache based on the to-be-fused product, and display the recommended product to the user according to the synchronized personalized product recommendation list.

[0099] Optionally, the data synchronization unit 40 can be specifically configured to, for any to-be-added product, insert the to-be-added product into the personalized product recommendation list at a position matched with a recommendation score of the to-be-added product, and save a unique product identifier and additional information of the to-be-added product into the local cache.

[0100] Optionally, the data synchronization unit 40 can be specifically configured to, for any to-be-updated product, compare real-time product key indicator data of the to-be-updated product in the customized product recommendation data with historical product key indicator data of the to-be-updated product in the preloaded product recommendation data, and if the real-time product key indicator data is better than the historical product key indicator data, replace the historical product key indicator data of the to-be-updated product in the personalized product recommendation list and the local cache with the real-time product key indicator data.

[0101] Optionally, the data synchronization unit 40 can be specifically configured to, for any to-be-deleted product, detect whether an activity level of the to-be-deleted product is lower than a preset activity threshold, and if so, delete the to-be-deleted product in the personalized product recommendation list and the local cache if the to-be-deleted product meets a preset deletion condition.

[0102] Optionally, the product personalized recommendation apparatus can further include a product segmentation unit, a thread allocation unit, a thread processing unit and a thread waiting unit.

[0103] The product segmentation subunit is configured to, in the process of data synchronization of the data synchronization unit 40 based on the to-be-fused product, divide each product into a plurality of product segmentation intervals according to a unique product identifier, and initialize a lock object for each product segmentation interval.

[0104] The thread allocation subunit is configured to start a plurality of threads, and assign each thread to process data synchronization of the to-be-fused product in one or more product segmentation intervals.

[0105] The thread processing subunit is configured to, when a thread is ready to start processing data synchronization of the to-be-fused product in a product segmentation interval, obtain the lock object if the lock object of the product segmentation interval is not currently occupied, start to traverse and process data synchronization of the to-be-fused product in the product segmentation interval, and release the lock object after the data synchronization of the to-be-fused product in the product segmentation interval.

[0106] The thread waiting subunit is configured to, when the lock object of the product segmentation interval is currently occupied, wait for the lock object to be released.

[0107] Optionally, the commodity personalized recommendation device can further include a data anomaly detection unit.

[0108] The data anomaly detection unit is configured to, based on the additional information of the to-be-fused commodities, detect whether there is data anomaly in the to-be-fused commodities in the process of the data synchronization unit 40 performing data synchronization on the personalized commodity recommendation list and the local cache based on the to-be-fused commodities, and in the case that any to-be-fused commodity is detected to have data anomaly, perform a rollback operation on the personalized commodity recommendation list and the local cache to restore the personalized commodity recommendation list and the local cache to a state before the present data synchronization.

[0109] Optionally, the commodity personalized recommendation device can further include a network monitoring unit.

[0110] The network monitoring unit is configured to, in the process of the data synchronization unit 40 performing data synchronization on the personalized commodity recommendation list and the local cache based on the to-be-fused commodities, monitor in real time whether there is network interruption in the data synchronization process, and if so, perform a rollback operation on the personalized commodity recommendation list and the local cache to restore the personalized commodity recommendation list and the local cache to a state before the present data synchronization.

[0111] Optionally, the rollback operation includes: in the order of the recommendation scores of the commodities in the preloaded commodity recommendation data from high to low before the present data synchronization, sequentially restoring the data in the personalized commodity recommendation list and the local cache, so that the preloaded commodity recommendation data in the restored personalized commodity recommendation list and the local cache is consistent with that before the present data synchronization.

[0112] Optionally, the commodity personalized recommendation device can further include a data validity detection unit.

[0113] The data validity detection unit is configured to detect the data validity of each commodity in the local cache, and in the case that any commodity has invalid data, trigger the cloud customized data obtaining unit 20.

[0114] The application provides a personalized commodity recommendation device, which is used for: in response to a commodity recommendation trigger operation of a user, asynchronously loading preloaded commodity recommendation data in a local cache in a personalized commodity recommendation list, wherein the preloaded commodity recommendation data comprises unique commodity identifiers and additional information of at least one commodity; uploading user behavior data and user preference parameters of the user to a cloud recommendation engine to obtain customized commodity recommendation data returned by the cloud recommendation engine, wherein the customized commodity recommendation data comprises unique commodity identifiers and additional information of at least one commodity; comparing the customized commodity recommendation data with the preloaded commodity recommendation data by using the unique commodity identifiers and the additional information to identify to-be-fused commodities, wherein the to-be-fused commodities comprise to-be-added commodities, to-be-updated commodities and / or to-be-deleted commodities; based on the to-be-fused commodities, performing data synchronization on the personalized commodity recommendation list and the local cache, and displaying recommended commodities to the user according to the synchronized personalized commodity recommendation list. Through real-time comparison and fusion of the locally preloaded data and the cloud customized recommendation data, the personalized commodity recommendation list is dynamically updated, fast response and accurate matching of commodity recommendation data are realized, and therefore the data loading and updating process of commodity recommendation is optimized, and the personalized effect and overall user experience of commodity recommendation are improved.

[0115] As to the device in the above-mentioned embodiments, the specific manner in which each unit performs operations has been described in detail in the embodiments related to the method, and will not be described in detail here.

[0116] The personalized commodity recommendation device comprises a processor and a memory, and the local data loading unit 10, the cloud customized data obtaining unit 20, the to-be-fused commodity identifying unit 30 and the data synchronization unit 40 are all stored in the memory as program units, and the corresponding functions are realized by the processor executing the program units stored in the memory.

[0117] The processor comprises a core, and the core retrieves the corresponding program units from the memory. One or more than one core can be set, and the real-time comparison and fusion of the cloud customized commodity recommendation data and the local preloaded commodity recommendation data are realized by adjusting the core parameters, the personalized commodity recommendation list is dynamically updated, the loading speed and matching accuracy of commodity recommendation data are improved, the commodity recommendation process is optimized, the personalized effect and overall user experience are enhanced.

[0118] The embodiment of the application provides a computer readable storage medium, which stores a program, and the program is executed by a processor to realize the personalized commodity recommendation method.

[0119] The embodiment of the application provides a processor, which is used for running a program, and the program is executed to realize the personalized commodity recommendation method.

[0120] AsFigure 3 As shown, the embodiment of the present application provides an electronic device 1000, which comprises at least one processor 1001, at least one memory 1002 connected with the processor 1001, and a bus 1003; wherein the processor 1001 and the memory 1002 complete mutual communication through the bus 1003; the processor 1001 is used to call program instructions in the memory 1002 to execute the above-mentioned product individualization recommendation method. The electronic device in the present application can be a server, a PC, a PAD, a mobile phone, etc.

[0121] The present application further provides a computer program product, which is suitable for executing the program initialized with the steps of the product individualization recommendation method when executed on an electronic device.

[0122] The present application is described with reference to flowcharts and / or block diagrams of the method, device, electronic device (system), and computer program product according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams can be realized by computer program instructions. These computer program instructions can be provided to a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable devices to produce a machine, so that the instructions executed by the computer or other programmable devices produce a device that implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The device that implements the functions specified in one flow or multiple flows and / or one block or multiple blocks. Figure 1 The device that implements the functions specified in one flow or multiple flows and / or one block or multiple blocks.

[0123] In a typical configuration, the electronic device comprises one or more processors (CPU), memory, and bus. The electronic device can further comprise an input / output interface, a network interface, etc.

[0124] The memory can comprise a non-permanent memory in a computer readable medium, random access memory (RAM) and / or non-volatile memory such as read-only memory (ROM) or flash memory (flash RAM), and the memory comprises at least one memory chip. The memory is an example of the computer readable medium.

[0125] Computer-readable media includes permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology for information storage. Information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette, magnetic disk storage or other magnetic storage device, or any other non-transmission medium that can be used to store information accessible by a computing device. According to the definition herein, computer-readable media does not include transitory media such as modulated data signals and carriers.

[0126] It should be noted that the user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of related data need to comply with relevant laws, regulations and standards of relevant countries and regions.

[0127] It can be understood that before using the technical solutions disclosed in the embodiments of the present disclosure, the type of personal information involved in the present disclosure, the use range, the use scene, etc. should be informed to the user and the authorization of the user should be obtained through appropriate means according to relevant laws and regulations.

[0128] In the description of the present application, it should be understood that if the terms "upper", "lower", "front", "rear", "left" and "right" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the position or element referred to must have a particular orientation, be constructed and operated in a particular orientation, therefore it cannot be understood as a limitation of the present application.

[0129] It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting; it is not intended to exclude myriad other embodiments of the present application that other present or future technologies can provide. It must be noted that as used herein, the singular forms "a", "an" and "the" include plural referents unless the context clearly dictates otherwise. The terms "includes" and "comprising," as well as derivatives thereof, mean that various embodiments include, but are not limited to, the listed material or steps. The terms "substantially," "essentially," "approximately," "about," or "perpendicular" and variations thereof, are defined as being close to as understood by one of ordinary skill in the art, and in one non-limiting embodiment the term "about" means within 10% of the stated value. The term "coupled" as used herein, means the joining of two members together such that the members together with the coupling are cohesively bound as a single integral piece.

[0130] Those skilled in the art will appreciate that embodiments of the present application can be devised for a method, a system, or a computer program product. Accordingly, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, etc.) embodying computer-readable program code.

[0131] The embodiments of the present application described above are merely given as examples of the present application. It is intended to cover any variations of the present application falling within the scope of the present application. Thus, the present application can have various modifications and various forms. It is also intended to cover one or more embodiments of the present application falling with the scope of the present application. Therefore, the scope of the present application should be limited only by the appended claims.

Claims

1. A product individualization recommendation method characterized by, The method comprises: in response to a user's commodity recommendation triggering operation, asynchronously loading preloaded commodity recommendation data in a local cache into a personalized commodity recommendation list, wherein the preloaded commodity recommendation data comprises unique commodity identifiers and additional information of at least one commodity; uploading user behavior data and user preference parameters of the user to a cloud recommendation engine to obtain customized commodity recommendation data returned by the cloud recommendation engine, wherein the customized commodity recommendation data comprises unique commodity identifiers and additional information of at least one commodity; comparing the customized commodity recommendation data with the preloaded commodity recommendation data by using the unique commodity identifiers and the additional information to identify to-be-fused commodities, wherein the to-be-fused commodities comprise to-be-added commodities, to-be-updated commodities and / or to-be-deleted commodities; synchronizing the to-be-fused commodities to the personalized commodity recommendation list and the local cache respectively, and displaying recommended commodities to the user according to the synchronized personalized commodity recommendation list.

2. The method of claim 1, wherein, The step of synchronizing the to-be-fused commodities to the personalized commodity recommendation list and the local cache respectively comprises: for any to-be-added commodity, inserting the to-be-added commodity into a position matched with a recommendation score of the to-be-added commodity in the personalized commodity recommendation list according to the recommendation score of the to-be-added commodity, and saving the unique commodity identifier and the additional information of the to-be-added commodity to the local cache; and / or, for any to-be-updated commodity, comparing real-time commodity key indicator data of the to-be-updated commodity in the customized commodity recommendation data with historical commodity key indicator data of the to-be-updated commodity in the preloaded commodity recommendation data, and if the real-time commodity key indicator data is better than the historical commodity key indicator data, replacing the historical commodity key indicator data in the personalized commodity recommendation list and the local cache with the real-time commodity key indicator data; and / or, for any to-be-deleted commodity, detecting whether the activity of the to-be-deleted commodity is lower than a preset activity threshold, and if so, deleting the to-be-deleted commodity in the personalized commodity recommendation list and the local cache if the to-be-deleted commodity meets a preset deletion condition.

3. The method of claim 1, wherein, In the process of synchronizing the to-be-fused commodities to the personalized commodity recommendation list and the local cache respectively, the method further comprises: dividing each commodity into a plurality of commodity segment intervals according to unique commodity identifiers, and initializing a lock object for each commodity segment interval; starting a plurality of threads, and assigning each thread to process data synchronization of the to-be-fused commodities in one or more commodity segment intervals; when a thread is ready to start processing data synchronization of the to-be-fused commodities in a commodity segment interval, if the lock object of the commodity segment interval is not currently occupied, obtaining the lock object, starting to traverse and process data synchronization of the to-be-fused commodities in the commodity segment interval, and releasing the lock object after the data synchronization of the to-be-fused commodities in the commodity segment interval; if the lock object of the commodity segment interval is currently occupied, the thread waits for the lock object to be released.

4. The method of claim 1, wherein, In the process of synchronizing the to-be-fused commodities to the personalized commodity recommendation list and the local cache respectively, the method further comprises: Based on the additional information of the to-be-fused commodities, it is detected whether there is data anomaly in the to-be-fused commodities, and in the case that it is detected that any of the to-be-fused commodities has data anomaly, a rollback operation is performed on the personalized commodity recommendation list and the local cache to restore the personalized commodity recommendation list and the local cache to the state before this time data synchronization.

5. The method of claim 1, wherein, In the process of synchronizing the to-be-fused commodities to the personalized commodity recommendation list and the local cache respectively, the method further comprises: Real-time monitoring whether network interruption occurs in the data synchronization process, and if so, performing a rollback operation on the personalized commodity recommendation list and the local cache to restore the personalized commodity recommendation list and the local cache to the state before this time data synchronization.

6. The method according to claim 4 or 5, characterized in that, The rollback operation comprises: In the order of the recommendation scores of each commodity in the preloaded commodity recommendation data from high to low before this time data synchronization, the data in the personalized commodity recommendation list and the local cache are restored in turn, so that the preloaded commodity recommendation data in the restored personalized commodity recommendation list and the local cache remain consistent with that before this time data synchronization.

7. The method of claim 1, wherein, Further comprising: Detecting the data validity of each commodity in the local cache, and in the case that any commodity has invalid data, performing the step of uploading the user behavior data and user preference parameters of the user to the cloud recommendation engine to obtain the customized commodity recommendation data returned by the cloud recommendation engine.

8. A commodity individualization recommendation apparatus characterized by comprising: Comprising: a local data loading unit, a cloud customized data obtaining unit, a to-be-fused commodity identifying unit and a data synchronization unit, The local data loading unit is configured to load preloaded commodity recommendation data in the local cache to a personalized commodity recommendation list asynchronously in response to a commodity recommendation triggering operation of a user, wherein the preloaded commodity recommendation data comprises a unique commodity identifier and additional information of at least one commodity. The cloud customized data obtaining unit is configured to upload user behavior data and user preference parameters of the user to a cloud recommendation engine to obtain customized commodity recommendation data returned by the cloud recommendation engine, wherein the customized commodity recommendation data comprises a unique commodity identifier and additional information of at least one commodity. The to-be-fused commodity identifying unit is configured to compare the customized commodity recommendation data with the preloaded commodity recommendation data by using the unique commodity identifier and the additional information to identify to-be-fused commodities, wherein the to-be-fused commodities comprise to-be-added commodities, to-be-updated commodities and / or to-be-deleted commodities. The data synchronization unit is configured to synchronize the to-be-fused commodities to the personalized commodity recommendation list and the local cache respectively, and display recommended commodities to the user according to the synchronized personalized commodity recommendation list.

9. A computer-readable storage medium having stored thereon a program, characterized in that, The program is executed by the processor to implement the commodity personalized recommendation method of any one of claims 1 to 7.

10. An electronic device, comprising: The electronic device comprises at least one processor, at least one memory connected with the processor, and a bus; wherein the processor, the memory and the bus complete mutual communication; the processor is used to call program instructions in the memory, so as to execute the commodity individualization recommendation method in any one of claims 1 to 7.

Images