Service recommendation method, apparatus, device, storage medium, and computer program product

Abstract

The application discloses a service recommendation method and device, equipment, a storage medium and a computer program product. The method comprises the following steps: when first service subscription information of a first user who subscribes to a service for the first time in a service system is acquired, a first service recommendation list of a second service having a subscription relationship with a first service in the first service subscription information is generated; a second service recommendation list including a third service is determined according to the first service subscription information and second service subscription information of a second user, the second user including at least one of the following: a family member of the first user, a resident member corresponding to the first user; the score information of each service in the second service and the third service is respectively adjusted through a correction parameter corresponding to the first service to obtain the score information of each service; and a target service in which the score information of the first service and the second service meets a preset score condition is recommended to the first user, so that the accuracy of recommending services to new users is improved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of computer, and particularly relates to a service recommendation method and device, equipment, a storage medium and a computer program product. BACKGROUND

[0002] With the continuous development of Internet technology, electronic devices can rely on application programs or short messages to recommend relevant services to users. In related technologies, the service system recommends services to new users by recommending the most popular services with the highest subscription in the service system to new users. However, since the popular services are not targeted, the service recommendation method has the problem of inaccuracy. SUMMARY

[0003] The embodiments of the present application provide a service recommendation method, device, equipment and storage medium, which can solve the problem of inaccurate service recommendation in related technologies.

[0004] In a first aspect, the embodiments of the present application provide a service recommendation method, which can include:

[0005] In the case of obtaining the first service subscription information of the first user, a first service recommendation list is generated based on the first service in the first service subscription information, the first user is a user who subscribes to a service for the first time in the service system, and the first service recommendation list includes a second service in the service system that has a subscription relationship with the first service;

[0006] According to the first service subscription information of the first user and the second service subscription information of the second user in the service system, a second service recommendation list is determined, the second user includes at least one of the following: a family member of the first user, a resident member corresponding to the first user, and the resident member includes at least one of the following: a daytime resident member and a nighttime resident member, and the second service recommendation list includes a third service in the second service subscription information;

[0007] The score information of each service in the second service and the third service is adjusted respectively by using the correction parameter corresponding to the first service, to obtain the score information of each service;

[0008] The target service is recommended to the first user, and the target service is a service in the first service and the second service whose score information meets a preset score condition.

[0009] In a second aspect, the embodiments of the present application provide a service recommendation device, which can include:

[0010] The generating module is configured to generate a first service recommendation list based on a first service in the first service subscription information of the first user, the first user being a user who subscribes to a service in the service system for the first time, and the first service recommendation list including a second service in the service system having a subscription relationship with the first service;

[0011] The determining module is configured to determine a second service recommendation list according to the first service subscription information of the first user and second service subscription information of a second user in the service system, the second user including at least one of the following: a family member of the first user, and a residence member corresponding to the first user, the residence member including at least one of the following: a daytime residence member and a nighttime residence member, and the second service recommendation list including a third service in the second service subscription information;

[0012] The adjusting module is configured to respectively adjust score information of each service in the second service and the third service by using a correction parameter corresponding to the first service, to obtain the score information of each service.

[0013] The recommending module is configured to recommend a target service to the first user, the target service being a service in the first service and the second service whose score information satisfies a preset score condition.

[0014] In a third aspect, an embodiment of the present application provides a computer device, which comprises a processor and a memory storing computer program instructions.

[0015] The processor executes the computer program instructions to implement the service recommendation method shown in the first aspect.

[0016] In a fourth aspect, an embodiment of the present application provides a computer storage medium, which stores computer program instructions, and the computer program instructions are executed by a processor to implement the service recommendation method shown in the first aspect.

[0017] In a fifth aspect, an embodiment of the present application provides a chip, which comprises a processor and a communication interface, the communication interface being coupled to the processor, and the processor being configured to run programs or instructions to implement the service recommendation method shown in the first aspect.

[0018] In a sixth aspect, an embodiment of the present application provides a computer program product stored in a storage medium, and the computer program product is executed by at least one processor to implement the service recommendation method shown in the first aspect.

[0019] The service recommendation method, device, equipment and storage medium provided by the embodiments of the present application, in the case of obtaining the first service subscription information of the first user, generate a first service recommendation list potentially recommended to the first user based on the first service in the first service subscription information, the first user is a user who subscribes to a service for the first time in a service system, and the first service recommendation list includes a second service in the service system having a subscription relationship with the first service. Then, according to the first service subscription information of the first user and the second service subscription information of a second user in the service system, determine a second service recommendation list potentially recommended to the first user in relation to the second user, the second user includes at least one of the following: a family member of the first user, a resident member corresponding to the first user, the resident member includes at least one of the following: a daytime resident member, a nighttime resident member, and the second service recommendation list includes a third service in the second service subscription information. Then, taking the effect of the first service actually handled by the user as a correction parameter, adjusting the score information of each service in the second service and the third service respectively to obtain the score information of each service, and recommending the target service after final revision to the first user, the target service is a service in the first service and the second service whose score information meets the preset score condition, in this way, the multiple types of services of the service system are comprehensively considered, so that the target service recommended to the first user has diversity, and since the first service subscribed by the user when entering the network is considered, the recommendation is more targeted than the recommendation based on popular services, the accuracy of recommending services to the first user is improved, at the same time, the first service recommendation list and the second service recommendation list effectively avoid the service rule that the same type of service cannot be handled repeatedly, effectively avoid homogeneity, and improve the efficiency of service recommendation. BRIEF DESCRIPTION OF DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments of the present application. Those skilled in the art can obtain other drawings according to these drawings without creating any creative labor.

[0021] Figure 1 A flowchart of a service recommendation method provided by the embodiments of the present application;

[0022] Figure 2 A first service recommendation list of a service recommendation method provided by the embodiments of the present application;

[0023] Figure 3 A second service recommendation list of a service recommendation method provided by the embodiments of the present application;

[0024] Figure 4 A service subscription relationship network diagram of a service recommendation method provided by the embodiments of the present application;

[0025] Figure 5 A schematic diagram of generating a target service of a service recommendation method provided by an embodiment of the present application;

[0026] Figure 6 A schematic diagram of determining a target service of a service recommendation method provided by an embodiment of the present application;

[0027] Figure 7 A structural schematic diagram of a service recommendation apparatus provided by an embodiment of the present application;

[0028] Figure 8 A structural schematic diagram of a computer device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0029] The features and exemplary embodiments of various aspects of the present application will be described below in detail, in order to make the purposes, technical solutions and advantages of the present application more clear and apparent, the present application will be further described in detail below in combination with the drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application, but not to limit the present application. The present application can be implemented without some of these specific details by those skilled in the art. The following description of the embodiments is only to provide a better understanding of the present application by showing examples of the present application.

[0030] It should be noted that, in this document, relational terms such as first and second and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any such actual relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without more limitations, an element defined by the statement "comprising" does not exclude the existence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0031] The acquisition, storage, use, processing, etc. of data (including but not limited to features, information, etc. in this document) in the technical solutions of the present application comply with the relevant provisions of national laws and regulations.

[0032] The service recommendation for new users is also called user cold start. The current service recommendation method can include the following three kinds, as shown below.

[0033] In a first mode, a score of an item is predicted according to score information of a user on a neighboring item, and the score is filled into a user-item score matrix to obtain a first filled score matrix; similarity between users is calculated according to the first filled score matrix; for each user, a user with high similarity is selected as a neighboring user of the user, and a score of an item not scored by the target user is predicted according to score information of the neighboring user, and the first filled score matrix is filled again; and the first N items with the highest predicted scores are selected from a final score matrix and recommended to the user. However, this service recommendation mode tends to repeatedly recommend services that have been ordered by the user to the user, and is homogeneous, and is difficult to expand product categories.

[0034] In a second mode, a user cold start service recommendation can be performed by using a local sensitive hashing algorithm. Service operation data in a current time window is collected, and a first popular service is selected. Service vector data is generated according to service self-provided information. Similarity between services is calculated based on the local sensitive hashing algorithm according to the service vector data. An association relationship between a service category and an interest label is established. If a user type of a new user is a tourist mode user, the first popular service is displayed as a first exposure service. If the user is a registered user, an interest label selected by the user is obtained, a second popular service is obtained, and a first exposure service is generated according to a preset weight and displayed. When the new user interacts with the first exposure service, an interactive similar service is found according to the similarity between services, and the similar service is displayed as a new exposure product. However, the popular service is not targeted, and the service recommendation mode has the problem of inaccuracy.

[0035] In a third mode, a user group attribute can be constructed by using a behavior footprint of a new user on other platforms. Then, according to the theory of people divided into groups, users are clustered and divided into multiple subsets according to user interest characteristics, and commodity preferences of users in each subset are calculated as a candidate set of users of this type. However, this mode needs to introduce external data, which increases the cost of service recommendation.

[0036] To solve the above technical problems, the embodiments of the present application provide a service recommendation method, device, computer equipment and storage medium.

[0037] Based on this, the service recommendation method, device, computer equipment and storage medium of the embodiments of the present application will be described in detail below with reference to the accompanying drawings. Figures 1 to 8 It should be noted that these embodiments are not intended to limit the scope of the present application.

[0038] Based on this, in order to better illustrate the content in the embodiments of the present application, the following will be described in combination with Figure 1 A service recommendation method provided by the embodiments of the present application is described as follows.

[0039] Figure 1 A flowchart of a service recommendation method provided for an embodiment of the present application.

[0040] As shown in Figure 1 The service recommendation method can be applied in a computer device, and the service recommendation method can specifically include the following steps:

[0041] In step 110, in a case where first service subscription information of a first user is acquired, a first service recommendation list is generated based on a first service in the first service subscription information, the first user is a user who subscribes to a service for the first time in a service system, and the first service recommendation list includes a second service in the service system that has a subscription relationship with the first service. In step 120, a second service recommendation list is determined according to the first service subscription information of the first user and second service subscription information of a second user in the service system, the second user includes at least one of the following: a family member of the first user, a resident member corresponding to the first user, and the resident member includes at least one of the following: a daytime resident member and a nighttime resident member, and the second service recommendation list includes a third service in the second service subscription information. In step 130, score information of each service in the second service and the third service is adjusted respectively by using a correction parameter corresponding to the first service, to obtain the score information of each service. In step 140, a target service is recommended to the first user, and the target service is a service in the first service and the second service whose score information satisfies a preset score condition.

[0042] In this way, in the case that the first service subscription information of the first user is acquired, a first service recommendation list potentially recommended to the first user is generated based on the first service in the first service subscription information, the first user is a user who subscribes to a service in the service system for the first time, and the first service recommendation list includes a second service in the service system having a subscription relationship with the first service. Then, according to the first service subscription information of the first user and second service subscription information of a second user in the service system, a second service recommendation list potentially recommended to the first user in relation to the second user is determined, the second user includes at least one of the following: a family member of the first user, a corresponding resident member of the first user, and the resident member includes at least one of the following: a daytime resident member and a nighttime resident member, and the second service recommendation list includes a third service in the second service subscription information. Then, taking the effect of the first service actually handled by the user as a correction parameter, the score information of each service in the second service and the third service is adjusted respectively to obtain the score information of each service, and the first user is recommended a target service after final revision, and the target service is a service in the first service and the second service whose score information meets a preset score condition. In this way, the multiple types of services of the service system are comprehensively considered, so that the target service recommended to the first user has diversity. Since the first service subscribed by the user when entering the network is considered, the recommendation is more targeted than the recommendation based on popular services, and the accuracy of the service recommended to the first user is improved. At the same time, the first service recommendation list and the second service recommendation list effectively avoid the service rule that the same type of service cannot be handled repeatedly, effectively avoid homogenization, and improve the efficiency of service recommendation.

[0043] The above steps are described in detail as follows.

[0044] Firstly, step 110 is involved. In some embodiments of the present application, the first service recommendation list can be acquired through service subscription relationship information stored in the service system. Based on this, the service system includes service subscription relationship information, and the step 110 can specifically include steps 1101 to 1104.

[0045] Step 1101: According to the first service subscription information, the fourth service having a subscription relationship with the first service is acquired from the service subscription relationship information, and the first promotion index corresponding to the first service is acquired. The first promotion index is used to represent the probability that the first service and the fourth service are subscribed by the user at the same time.

[0046] Exemplarily, the service system determines the user who subscribes to the service in the service system for the first time as the first user, determines the service subscribed by the first user as the first service, and determines the information of the first service as the first service subscription information, in the case of detecting the behavior of the user subscribing to the service in the service system for the first time. In this way, the fourth service such as the fourth service B and the fourth service C having a subscription relationship with the first service A subscribed by the first user for the first time is searched from the service subscription relationship information stored in the service system according to the first service A. At this time, the number of the first promotion indexes corresponding to the first service A corresponds to the number of the fourth services, so that the first promotion indexes include the first promotion index (A-B) and the first promotion index (A-C) in this example, where the first promotion index (A-B) is used to represent the probability that the first service A and the fourth service B are subscribed by the user at the same time, for example, 1.75, and the first promotion index (A-C) is used to represent the probability that the first service A and the fourth service C are subscribed by the user at the same time, for example, 1.5.

[0047] Step 1102, screening the second promotion index greater than or equal to the preset promotion index from the first promotion index.

[0048] Exemplarily, if the preset promotion index is 1, the first promotion index (A-B) and the first promotion index (A-C) can be determined as the second promotion index.

[0049] Step 1103, determining the service corresponding to the second promotion index as the second service.

[0050] Exemplarily, the fourth service B and the fourth service C can be determined as the second service.

[0051] Step 1104, arranging the second service in the order of the promotion index from high to low to obtain the first service recommendation list.

[0052] Exemplarily, since the first promotion index (A-B) is higher than the first promotion index (A-C), the fourth service B is arranged before the fourth service C to obtain the first service recommendation list (B and C).

[0053] It should be noted that the first service can be a single service in the service system, or a collection of multiple single services.

[0054] Here, the embodiment of the present application provides a step of generating the above-mentioned service subscription relationship information, which is used to collect the service subscription information of all users subscribing to the service in the service system, so as to generate the first service recommendation list through the service subscription relationship information in the user dimension.

[0055] Based on this, before step 110, the service recommendation method can further include steps 1501 to 1503.

[0056] In step 1501, third service subscription information of N users of a service system in a first preset time window is acquired, the third service subscription information including M services subscribed by the N users in the service system, N being a positive integer and M being an integer greater than 1.

[0057] Exemplarily, since the service subscription period is long, in order to acquire relatively stable services used by users, half a year can be selected as the time dimension of service subscription information extraction, i.e., the first preset time window. Based on this, third service subscription information of services subscribed in the service system in half a year can be selected, as shown in Table 1. Figure 2 As shown in Table 1, N is 3 and M is 3, i.e., taking user 10, user 20 and user 30 as examples, the services subscribed by user 10 in half a year are service 1 and service 2, the services subscribed by user 20 are service 1 and service 3, and the services subscribed by user 30 are service 1, service 2 and service 3.

[0058] In step 1502, according to the third service subscription information, the lift index of the i-th service is calculated by using an association rule algorithm, the lift index of the i-th service being used to represent the probability that the i-th service and the j-th service in the M services are simultaneously subscribed by a user, i∈[1, M], j∈[1, M], i≠j.

[0059] Exemplarily, for the subscription relationship between users and services as shown in Table 1, the support index and the execution index of each service involved by user 10, user 20 and user 30 are calculated by using the association rule algorithm, so as to calculate the lift index based on the support index and the execution index of the service, and to obtain service subscription relationship information according to the lift index of the service. Figure 2

[0060] Specifically, in the embodiment of the present application, the lift index of the i-th service can be calculated based on the support index of the i-th service and the confidence index of the i-th service, i.e., step 1502 can specifically include steps 15021 to 15023.

[0061] In step 15021, according to the first service set to which the i-th service belongs and P service sets in the service system, the support index of the i-th service is calculated by using the association rule algorithm, the support index of the i-th service being used to represent the proportion of the first service set including the i-th service in the P service sets, the P service sets including the first service set.

[0062] Here, the support index can represent the proportion of the service set containing service 1 in all service sets in the service system, which can be understood as the probability of service selection.

[0063] ​Exemplarily, still taking the example in the above steps, taking the user 10, the user 20 and the user 30 as examples, i can be 1, 2 and 3, the first service in the three services, that is, the service 1, the first service set including the service 1 can be 2, P is 10, at this time, the support index of the service 1 is 2 / 10 = 0.2. Similarly, the first service set of the service 2 can be 4, at this time, the support index of the service 2 is 4 / 10 = 0.4. Similarly, the first service set of the service 3 can be 5, at this time, the support index of the service 3 is 5 / 10 = 0.5.

[0064] In step 15022, according to the first service set to which the i-th service belongs and the fifth service in the first service set except the i-th service, the confidence index of the i-th service is calculated by using the association rule algorithm, and the confidence index is used to represent the proportion of the existence of the fifth service in the first service set except the i-th service.

[0065] Exemplarily, still taking the example in the above steps as an example, the confidence index is a conditional probability, without considering the order of subscription, two conditional probabilities are calculated at the same time, for example, the probability that the service 1 belongs to the first service set 1 including the service 2, combined with the service 2, the service 3 and the service 4 in the first service set 1 of the service 1, as shown in the following table: Figure 2 As shown in the table, each user's handled service is regarded as a service set, then the first service set of the service 1 has three, the service set including the service 1 and the service 2 has two, accounting for 2 / 3 = 0.67 of the total three service sets.

[0066] It should be noted that the association rule algorithm in the embodiment of the application can include at least one of the following: the Apriori algorithm, the FPGrowth algorithm. The Apriori algorithm and the FPGrowth algorithm in the application are widely used in the mining processing of the data in the value-added business of the operator. The PrefixSpan algorithm well makes up for the small adaptability of the Apriori algorithm or the FPGrowth algorithm, relieves the pressure of a large number of intermediate item sets (that is, sequences) generated by excessive scanning of the database, improves the mining efficiency, and is a frequently used improved algorithm in the engineering scene of the operator.

[0067] In addition, the order of handling can also be considered, and the association rule algorithm can also be implemented using improved algorithms such as PrefixSpan, i.e., the Prefix-Projected Pattern Growth (PrefixSpa) algorithm, which can solve the problem of mining frequent sequences that meet the minimum support from prefixes with a length of 1 recursively until the frequent sequences corresponding to longer prefixes cannot be mined. In this process, the PrefixSpan algorithm has the characteristics of stability and does not generate too many candidate sequences, so that the projected database can quickly consume memory and does not bring too much running pressure to the database.

[0068] In step 15023, the ratio of the confidence index and the support index is determined as the lift index of the ith service.

[0069] Exemplarily, the lift index is an important measure for generating the first service recommendation list, and the lift index is the frequency of the simultaneous occurrence of one service and another service, which is related to the frequency of the occurrence of one service and another service. Taking the lift index of service 1 as an example, the ratio of the confidence index 0.67 of service 1 and the support index 0.2 of service 1, i.e., 3.35, is determined as the lift index of service 1.

[0070] Based on this, when the lift index is greater than 1, the service is recommended, and when the lift index is less than 1, the service is not recommended. For example, lift index > 1: representing promotion, which can be recommended; lift index = 1: representing no promotion or no decline; and lift index < 1: representing decline, which is not recommended.

[0071] In step 1503, the ith service, the jth service, and the lift index of the ith service are associated to obtain service subscription relationship information.

[0072] Exemplarily, service 1 with a lift index greater than or equal to 1, service 2 related to service 1, and the lift index of service 1 are associated to obtain service subscription relationship information. Here, the information can be stored in a dictionary format, with service 1 as the key, service 2 as the service having a subscription relationship under the condition of the probability of service 1, and the lift index is also recorded for subsequent sorting.

[0073] Next, step 120 is involved, and based on the home address, daytime residence, and nighttime residence of the first user, the service subscription information of the family members, daytime residence members, and nighttime residence members in the service system is searched respectively to generate a second service recommendation list.

[0074] Based on this, in some embodiments of the present application, the step 120 can specifically include steps 1201 to 1204, which are specifically as follows.

[0075] At step 1201, a user-service subscription relationship network graph is constructed according to the first service subscription information of the first user and the second service subscription information of the second user in the service system, wherein the user-service subscription relationship network graph comprises first nodes, second nodes and connecting edges connecting the first nodes and the second nodes, the first nodes comprise first sub-nodes and second sub-nodes, the first sub-nodes are nodes corresponding to the first user, the second sub-nodes are nodes corresponding to the second user, the second nodes correspond to the third service, and the connecting edges are used to represent the subscription relationship between the user and the service.

[0076] As shown in the example of FIG. 6, the number of the second users can be four. The family members can be user1 and user2, the daytime resident member user3 and the nighttime resident member user4. The right-hand service of each second user is the second service subscription information of the second user. Based on the second service subscription information as shown in FIG. 6, the user-service subscription relationship network graph is constructed by establishing the undirected link between all the services and the users having the subscription relationship. Figure 3 Figure 3 As shown in the example of FIG. 6, the number of the second users can be four. The family members can be user1 and user2, the daytime resident member user3 and the nighttime resident member user4. The right-hand service of each second user is the second service subscription information of the second user. Based on the second service subscription information as shown in FIG. 6, the user-service subscription relationship network graph is constructed by establishing the undirected link between all the services and the users having the subscription relationship. Figure 4

[0077] Here, it should be noted that the reason for establishing the user-service subscription relationship network graph, i.e., the undirected link, is to facilitate the subsequent random walk. The user-service subscription relationship network graph can be constructed by using the related tool, such as Networks, and the recommendation algorithm PersonalRank of the graph.

[0078] At step 1202, the random walk is performed in the user-service subscription relationship network graph with a preset probability starting from the first sub-node, and the node influence index value of the second node is calculated. The node influence index value is the sum of the first probability value of randomly selecting the second node from the starting point and the second probability value of randomly walking from the second node to the starting point.

[0079] As an example, for the user-service subscription relationship network graph, the walk is started from the node corresponding to the first user u with a probability d, and the walk is restarted from the node corresponding to the first user u with a probability (1-d). When the node influence index value of the node corresponding to the service tends to be stable, the node influence index value is obtained.

[0080] The node influence index value of the second node can be calculated by the following formula (1):

[0081]

[0082] wherein Vi represents one of the second nodes, d represents a damping factor, i.e., a probability, Vj is a node connected to Vi, (1-d) represents a probability of restarting the walk from the node corresponding to the first user u, and r represents a probability of randomly walking from the node corresponding to the first user u to the node corresponding to the second user. i ​​a first probability value for the user to randomly select a second type node Vi from a starting point in the network diagram of the user-service subscription relationship, a second probability value for the random walk from Vi to the starting point, and ri is a prior random vector representing a case of randomly selecting a node.

[0083] Step 1203: filtering, from the node influence index values, a target node influence index value greater than or equal to a preset node influence index value, the target node influence index value corresponding to a target node in the second type nodes.

[0084] Exemplarily, in order to consider the risk of downshifting, a service of the same type as the service subscribed by the first user and with a lower cost can be excluded, the node influence index values calculated in the last step can be arranged in descending order, and the target node influence index value can be filtered according to the preset node influence index value.

[0085] Step 1204: generating a second service recommendation list according to a target service corresponding to the target node.

[0086] It should be noted that the embodiments of the present application also provide a manner of determining second service subscription information according to different user types of second users, which is shown as follows.

[0087] In some embodiments of the present application, the second user is a family member, and the second service subscription information includes first target service subscription information. Based on this, before step 120, the service recommendation method can further include steps 1601 and 1602.

[0088] Step 1601: obtaining, based on first home address information of the first user, second home address information matched with the first home address information from a service system.

[0089] Step 1602: determining a user corresponding to the second home address information as a family member, and determining service subscription information of the family member in the service system as the first target service subscription information.

[0090] In some other embodiments of the present application, the second user is a resident member, and the second service subscription information includes second target service subscription information. Based on this, before step 120, the service recommendation method can further include steps 1603 and 1604.

[0091] Step 1603: obtaining, based on device information of the first user, operator base station data corresponding to the first user, the operator base station data including positioning data of the first user.

[0092] Step 1604: determining a user in an area represented by the positioning data as a resident member, and determining service subscription information of the resident member in the service system as the second target service subscription information.

[0093] It should be noted that, since there are many daytime and nighttime station members, a fixed number of online users can be extracted using random sampling.

[0094] Furthermore, regarding step 130, in some embodiments of this application, the service recommendation method may also include steps 1701 and 1702 prior to step 130.

[0095] Step 1701: Calculate at least two third probability values ​​for the first service that were ordered by users in the service system within the second preset time window.

[0096] Step 1702: Normalize at least two third probability values ​​to obtain the correction parameters corresponding to the first service.

[0097] For example, the correction parameter w can be the result of normalizing the historical processing probability using the statistical method.

[0098] Based on this, step 130 will be explained in detail through the following example.

[0099] like Figure 5 As shown, the first service recommendation list generated by combining the above steps is recommendation list1, and the second service recommendation list includes a family network list list2 generated from the process of finding family members by home address, a daytime residence network list list3 generated from daytime resident members, and a nighttime residence network list list4 generated from nighttime resident members, generating a list as follows. Figure 6 The content shown is then used to adjust the scores of each service (service 1, p1; service 2, p2; service 3, p3; service 4, p4; and service 5, p5) with the correction parameter w to obtain the target service.

[0100] The service rating information can be calculated using the following formula (2):

[0101] Score(P j )=∑W i *P ij (2)

[0102] Where Pj represents the stable information of the service, and Pij represents the j-th product in the i-th list.

[0103] Then, regarding step 140, since users have diverse interests and the overall ranking may contain products of certain categories, after obtaining the overall calculation result, a certain number of products will be appropriately rearranged to consider factors such as diversity. Based on this, in some embodiments of this application, before step 140, the service recommendation method may also include steps 1801 and 1802.

[0104] In step 1801, a first probability distribution of a first service type of the target service in service types in the service system and a second probability distribution of a first service resource amount of the target service in service resource amounts in the service system are obtained.

[0105] For example, the first probability distribution p = (p1…p k ) of the first service type of the target service in the service types in the service system and the second probability distribution Q = (q1…q k ) of the first service resource amount of the target service in the service resource amounts in the service system are obtained. For the operator field, there are mainly two diversities of services, price diversity and service category diversity. The price is classified in certain intervals, such as 0-50 yuan, 51-100 yuan, etc.; the service categories include main packages, value-added packages, rights products, broadband products and peripheral products, etc.

[0106] It should be noted that the first probability distribution and the second probability distribution in the embodiment of the application are both normal distributions of price and product category. In step 1802, the similarity of the first probability distribution and the second probability distribution is calculated by the Hellinger distance algorithm to obtain a Hellinger distance value, which is used to represent the diversity distribution degree of the target service in the service system.

[0107] For example, the Hellinger distance value between the diversity distribution of the target service and the ideal diversity distribution is calculated to measure the degree of diversity of the target service.

[0108] The Hellinger distance value (Hellinger) can be calculated by the following formula (3):

[0109]

[0110] Where i is the target service, and k is the number of target services. Based on this, in this embodiment, the step 140 can specifically include:

[0111] In the case where the Hellinger distance value is greater than or equal to a preset distance value, the target service is recommended to the first user.

[0112] In addition, before step 140, the service recommendation method can further include:

[0113] In the case where the Hellinger distance value is less than a preset distance value, a second service type other than the first service type is obtained from the service system.

[0114] According to the sixth service corresponding to the second service type, the target service is updated; wherein the way of updating the target service can be replacing the end service in the target service according to the missing type, or directly adding the sixth service to the determined target service.

[0115] Based on this, the step 140 can specifically include:

[0116] The first user is recommended with the updated target service.

[0117] Therefore, the embodiment of the present application can generate a potential first service recommendation list according to the association rule based on the package or product of the first user entering the network, and construct a service subscription relationship network graph based on the service subscription information of the second user related to the first user. Based on the first sub-node corresponding to the first user as the starting point, the PersonalRank algorithm is used to perform random walk in the user and service subscription relationship network graph with a preset probability, and the node influence index value of the second type node is calculated. In this way, since the first user cannot subscribe to the same category of services at the same time, the same category of services will not be bound in the second service recommendation list, and the phenomenon of repeated recommendation of the same category is avoided. At the same time, since the PersonalRank algorithm is used for recommendation, the second service recommendation list will have a certain diversity, and the generation of the first service recommendation list and the second service recommendation list needs to consider the risk of downgrading, that is, the Hellinger distance between the diversity distribution of the generated target service and the ideal diversity distribution is calculated to measure the goodness of the result diversity of the target service, and finally the service is updated according to the result.

[0118] The service recommendation method in the embodiment of the present application considers the first service subscribed by the user when entering the network, which is more personalized than the popular product, and does not need to collect interest tags when the user enters the network. In addition, the present application does not need to access external data, has a great cost advantage, and reduces the cost of service recommendation. Moreover, the recommended service recommendation list effectively avoids the business rules that the same type of service cannot be repeated, and secondly, the recommended service recommendation list effectively avoids homogenization and improves the efficiency of service recommendation.

[0119] The present application also provides a service recommendation device, which will be described in detail in combination with Figure 7 .

[0120] Figure 7 is a structural schematic diagram of the service recommendation device provided by an embodiment of the present application.

[0121] In some embodiments of the present application, Figure 7 The service recommendation device shown can be provided in the computing device provided by the embodiment of the present application.

[0122] As Figure 7As shown, the service recommendation apparatus 70 can specifically include:

[0123] a generating module 701 configured to, in a case where the first service subscription information of the first user is acquired, generate a first service recommendation list based on a first service in the first service subscription information, the first user being a user who subscribes to a service in the service system for the first time, the first service recommendation list including a second service in the service system having a subscription relationship with the first service;

[0124] a determining module 702 configured to determine a second service recommendation list according to the first service subscription information of the first user and second service subscription information of a second user in the service system, the second user including at least one of the following: a family member of the first user, a corresponding resident member of the first user, the resident member including at least one of the following: a daytime resident member, a nighttime resident member, the second service recommendation list including a third service in the second service subscription information;

[0125] an adjusting module 703 configured to respectively adjust score information of each service in the second service and the third service by using a correction parameter corresponding to the first service, to obtain the score information of each service;

[0126] a recommendation module 704 configured to recommend a target service to the first user, the target service being a service in the first service and the second service whose score information satisfies a preset score condition.

[0127] In this way, in the case that the first service subscription information of the first user is acquired, a first service recommendation list potentially recommended to the first user is generated based on the first service in the first service subscription information, the first user is a user who subscribes to a service for the first time in the service system, and the first service recommendation list includes a second service in the service system having a subscription relationship with the first service. Then, according to the first service subscription information of the first user and second service subscription information of a second user in the service system, a second service recommendation list potentially recommended to the first user in relation to the second user is determined, the second user includes at least one of the following: a family member of the first user, a corresponding resident member of the first user, and the resident member includes at least one of the following: a daytime resident member and a nighttime resident member, and the second service recommendation list includes a third service in the second service subscription information. Then, taking the effect generated by the first service actually handled by the user as a correction parameter, the score information of each service in the second service and the third service is adjusted respectively to obtain the score information of each service, and the target service after final revision is recommended to the first user, the target service is a service in the first service and the second service whose score information meets a preset score condition. In this way, the multiple types of services in the service system are comprehensively considered, so that the target service recommended to the first user has diversity. Since the first service subscribed by the user when entering the network is considered, the recommendation is more targeted than the recommendation based on popular services, the accuracy of recommending services to the first user is improved, and at the same time, the first service recommendation list and the second service recommendation list effectively avoid the service rule that the same type of service cannot be handled repeatedly, effectively avoid homogenization, and improve the efficiency of service recommendation.

[0128] The service recommendation device 70 in the embodiments of the present application will be described in detail below.

[0129] In some embodiments of the present application, the generation module 701 can be specifically configured to, in the case that the service system includes service subscription relationship information, acquire, according to the first service subscription information, a fourth service having a subscription relationship with the first service from the service subscription relationship information and a first promotion index corresponding to the first service, the first promotion index being used to represent the probability that the first service and the fourth service are subscribed by the user at the same time; and filter a second promotion index greater than or equal to a preset promotion index from the first promotion index.

[0130] The service corresponding to the second promotion index is determined as the second service.

[0131] The second service is arranged in descending order of the promotion index to obtain the first service recommendation list.

[0132] In some other embodiments of the present application, the service recommendation device 70 in the embodiments of the present application can further include a first acquisition module, a first calculation module and an association module; wherein,

[0133] The first obtaining module is configured to obtain third service subscription information of N users of a service system within a first preset time window, the third service subscription information comprising M services subscribed by the N users in the service system, N being a positive integer, and M being an integer greater than 1;

[0134] The first calculating module is configured to calculate, according to the third service subscription information, a lift index of an i th service by using an association rule algorithm, the lift index of the i th service being used to represent a probability that the i th service and a j th service are simultaneously subscribed by a user in the M services, i ∈ [1, M], j ∈ [1, M], and i ≠ j;

[0135] The associating module is configured to associate the i th service, the j th service, and the lift index of the i th service to obtain service subscription relationship information.

[0136] In some other embodiments of the present application, the first calculating module in the embodiments of the present application can be specifically configured to calculate, according to a first service set to which the i th service belongs and P service sets in the service system, a support index of the i th service by using an association rule algorithm, the support index of the i th service being used to represent a proportion of the first service set to the P service sets, the P service sets including the first service set;

[0137] The confidence index is used to represent a proportion of the first service set in which the fifth service exists in addition to the i th service.

[0138] The ratio of the confidence index to the support index is determined as the lift index of the i th service.

[0139] In some other embodiments of the present application, the determining module 702 in the embodiments of the present application can be specifically configured to construct a user-service subscription relationship network graph according to the first service subscription information of the first user and second service subscription information of a second user in the service system, wherein the user-service subscription relationship network graph comprises first-type nodes, second-type nodes, and connecting edges connecting the first-type nodes and the second-type nodes, the first-type nodes comprise first sub-nodes and second sub-nodes, the first sub-nodes are nodes corresponding to the first user, the second sub-nodes are nodes corresponding to the second user, the second-type nodes correspond to third services, and the connecting edges are used to represent subscription relationships between users and services.

[0140] The first sub-nodes are taken as starting points, random walks are performed in the user-service subscription relationship network graph at a preset probability, and a node influence index value of the second-type nodes is calculated, the node influence index value being a sum of a first probability value of randomly selecting the second-type nodes from the starting points and a second probability value of randomly walking from the second-type nodes to the starting points.

[0141] filtering, from the node influence index values, a target node influence index value greater than or equal to a preset node influence index value, the target node influence index value corresponding to a target node in the second type of nodes;

[0142] generating a second service recommendation list according to a target service corresponding to the target node.

[0143] In some other embodiments of the present application, the service recommendation apparatus 70 in the embodiments of the present application can further include a second acquisition module; wherein,

[0144] The second acquisition module is configured to, in a case where the second user is a family member and the second service subscription information includes the first target service subscription information, acquire, based on the first family address information of the first user, second family address information matched with the first family address information from the service system.

[0145] The determination module 702 can be further configured to determine a user corresponding to the second family address information as a family member, and determine service subscription information of the family member in the service system as the first target service subscription information.

[0146] In some other embodiments of the present application, the service recommendation apparatus 70 in the embodiments of the present application can further include a third acquisition module; wherein,

[0147] The third acquisition module is configured to, in a case where the second user is a resident member and the second service subscription information includes second target service subscription information, acquire, based on the equipment information of the first user, operator base station data corresponding to the first user, the operator base station data including positioning data of the first user.

[0148] The determination module 702 can be further configured to determine a user within an area represented by the positioning data as a resident member, and determine service subscription information of the resident member in the service system as the second target service subscription information.

[0149] In some other embodiments of the present application, the service recommendation apparatus 70 in the embodiments of the present application can further include a statistical module and a processing module; wherein,

[0150] The statistical module is configured to statistically acquire at least two third probability values of the first service being subscribed by users in the service system within a second preset time window.

[0151] The processing module is configured to perform normalization processing on the at least two third probability values to obtain a correction parameter corresponding to the first service.

[0152] In some other embodiments of the present application, the service recommendation apparatus 70 in the embodiments of the present application can further include a third acquisition module and a second calculation module; wherein,

[0153] a third obtaining module, configured to obtain a first probability distribution of a first service type of the target service in service types in the service system and a second probability distribution of a first service resource amount of the target service in service resource amounts in the service system;

[0154] a second calculating module, configured to calculate a similarity of the first probability distribution and the second probability distribution by using a Hellinger distance algorithm to obtain a Hellinger distance value, the Hellinger distance value being used to represent a diversity distribution degree of the target service in the service system;

[0155] The recommendation module 704 can be specifically configured to recommend the target service to the first user in a case where the Hellinger distance value is greater than or equal to a preset distance value.

[0156] In still some embodiments of the present application, the service recommendation apparatus 70 in the embodiments of the present application can further include a fourth obtaining module and an updating module; wherein,

[0157] The fourth obtaining module is configured to obtain, in a case where the Hellinger distance value is less than the preset distance value, a second service type other than the first service type from the service system.

[0158] The updating module is configured to update the target service according to a sixth service corresponding to the second service type.

[0159] The recommendation module 704 can be specifically configured to recommend the updated target service to the first user.

[0160] Based on the same inventive concept, the present application further provides a computer device. The computer device will be specifically described in combination with Figure 8 the above technical scheme.

[0161] Figure 8 FIG. 1 is a structural schematic diagram of a computer device provided by an embodiment of the present application.

[0162] As Figure 8 shown, the computer device can include at least one of the following: an electronic device, a server, involved in the embodiments of the present application. The computer device can include a processor 801 and a memory 802 having computer program instructions stored therein.

[0163] Specifically, the processor 801 can include a central processing unit (CPU), or a specific integrated circuit (Application Specific Integrated Circuit, ASIC), or can be configured to implement one or more integrated circuits of the embodiments of the present application.

[0164] The memory 802 can include mass storage for data or instructions. As an example and not by way of limitation, the memory 802 can include a hard disk drive (HDD), floppy disk drive, flash memory, compact disk (CD), digital versatile disk (DVD), optical disk, tape, universal serial bus (USB) drive, or combinations of two or more of these. Where appropriate, the memory 802 can include removable or non-removable (or fixed) media. Where appropriate, the memory 802 can be internal or external to the integrated gateway disaster recovery appliance. In particular embodiments, the memory 802 is non-volatile, solid-state memory. In particular embodiments, the memory 802 includes read-only memory (ROM). Where appropriate, this ROM can be mask-programmed ROM, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), electrically alterable ROM (EAROM), or flash memory or combinations of two or more of these.

[0165] The processor 801 implements the service recommendation method of any of the above-described embodiments by reading and executing computer program instructions stored in the memory 802.

[0166] In one example, the computer device can also include a communication interface 803 and a bus 810. As shown, the processor 801, the memory 802, the communication interface 803 are connected through the bus 810 and complete the communication between each other. Figure 8

[0167] The communication interface 803 is mainly used to realize the communication between the modules, devices, units and / or equipment in the embodiments of the application.

[0168] The bus 810 includes hardware, software, or both, that couples components of the traffic control device to each other in a communicative manner. As an example and not by way of limitation, the bus can include an accelerated graphics port (AGP) or other graphics bus, an enhanced industry standard architecture (EISA) bus, a front-side bus (FSB), a HyperTransport (HT) interconnect, an industry standard architecture (ISA) bus, an InfiniBand (IB) interconnect, a low pin count (LPC) bus, a memory bus, a Micro Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a serial advanced technology attachment (SATA) bus, a video electronics standards association local (VLB) bus, or another suitable bus or interconnect, or combinations of two or more of these. Where appropriate, the bus 810 can include one or more buses. Although the present application describes and illustrates a particular bus, the present application contemplates any suitable bus or interconnect.

[0169] ​The service recommendation device can execute the service recommendation method in the embodiments of the present application, thereby realizing the service recommendation method and device described in combination Figures 1 to 8 with the service recommendation method in the embodiments described above.

[0170] In addition, in combination with the service recommendation method in the embodiments described above, the embodiments of the present application can provide a computer readable storage medium for implementation. The computer readable storage medium has computer program instructions stored thereon; the computer program instructions are executed by a processor to implement any one of the service recommendation methods in the embodiments described above.

[0171] It needs to be made clear that the present application is not limited to the specific configurations and processes described above and shown in the drawings. For the sake of brevity, detailed descriptions of well-known methods are omitted herein. In the embodiments described above, several specific steps are described and shown as examples. However, the method processes of the present application are not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order of the steps, after understanding the spirit of the present application.

[0172] The functional blocks shown in the structural block diagrams above can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a functional card, etc. When implemented in software, the elements of the present application are program or code segments used to perform the required tasks. The program or code segments can be stored in a machine readable medium or transmitted through a data signal carried in a carrier wave over a transmission medium or communication link. The "machine readable medium" can include any medium capable of storing or transmitting information. Examples of the machine readable medium include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, etc. The code segments can be downloaded via a computer network such as the Internet, an intranet, etc.

[0173] It also needs to be made clear that the exemplary embodiments mentioned in the present application describe some methods or systems based on a series of steps or devices. However, the present application is not limited to the order of the steps described above, that is, the steps can be executed in the order mentioned in the embodiments, or in an order different from the embodiments, or several steps can be executed simultaneously.

[0174] The above is merely a specific implementation of the present application. As can be clearly understood by a person skilled in the art from the above description, for the convenience and brevity of description, the specific working process of the system, module and unit described above can refer to the corresponding process in the foregoing method embodiments, which will not be described herein again. It should be understood that the protection scope of the present application is not limited in this way, and any person skilled in the art can easily think of various equivalent modifications or replacements within the technical range disclosed in the present application, and these modifications or replacements should be covered in the protection scope of the present application.

Claims

1. A service recommendation method characterized by comprising: The method comprises the following steps: In the case that the first service subscription information of the first user is acquired, a first service recommendation list is generated based on the first service in the first service subscription information, the first user is a user who subscribes to a service for the first time in a service system, and the first service recommendation list comprises a second service in the service system which has a subscription relationship with the first service; A second service recommendation list is determined according to the first service subscription information of the first user and the second service subscription information of a second user in the service system, the second user comprises at least one of the following: a family member of the first user, a residence member corresponding to the first user, the residence member comprises at least one of the following: a daytime residence member and a nighttime residence member, and the second service recommendation list comprises a third service in the second service subscription information; The score information of each service in the second service and the third service is respectively adjusted by a correction parameter corresponding to the first service, so as to obtain the score information of each service; A target service is recommended to the first user, the target service is a service in the first service and the second service whose score information satisfies a preset score condition.

2. The method of claim 1, wherein, The service system comprises service subscription relationship information; The first service recommendation list is generated based on the first service in the first service subscription information, which comprises the following steps: According to the first service subscription information, a fourth service having a subscription relationship with the first service and a first promotion index corresponding to the first service are acquired from the service subscription relationship information, and the first promotion index is used to represent the probability that the first service and the fourth service are subscribed by a user at the same time; Second promotion indexes greater than or equal to a preset promotion index are screened from the first promotion index; Services corresponding to the second promotion indexes are determined as the second service; The second service is arranged in descending order of promotion index, so as to obtain the first service recommendation list.

3. The method of claim 1, wherein, The second service recommendation list is determined according to the first service subscription information of the first user and the second service subscription information of the second user in the service system, which comprises the following steps: According to the first service subscription information of the first user and the second service subscription information of the second user in the service system, a user-service subscription relationship network graph is constructed, wherein the service subscription relationship network graph comprises a first type of node, a second type of node and a connection edge connecting the first type of node and the second type of node, the first type of node comprises a first sub-node and a second sub-node, the first sub-node is a node corresponding to the first user, the second sub-node is a node corresponding to the second user, the second type of node corresponds to the third service, and the connection edge is used to represent the subscription relationship between a user and a service. starting from the first sub-node, performing random walk in the user-service subscription relationship network graph with a preset probability, and calculating a node influence index value of the second type of node, the node influence index value being a sum of a first probability value of randomly selecting from the starting point to the second type of node and a second probability value of randomly walking from the second type of node to the starting point; filtering a target node influence index value greater than or equal to a preset node influence index value from the node influence index value, the target node influence index value corresponding to a target node in the second type of node; generating the second service recommendation list according to a target service corresponding to the target node.

4. The method of claim 1, wherein, The second user is the family member, and the second service subscription information includes first target service subscription information; Before the determining the second service recommendation list according to the second service subscription information of the second user in the service system, the method further includes: obtaining second home address information matched with the first home address information of the first user from the service system based on the first home address information of the first user; determining a user corresponding to the second home address information as the family member, and determining service subscription information of the family member in the service system as the first target service subscription information.

5. The method of claim 1, wherein, The second user is the resident member, and the second service subscription information includes second target service subscription information; Before the determining the second service recommendation list according to the second service subscription information of the second user in the service system, the method further includes: obtaining operator base station data corresponding to the first user based on the device information of the first user, the operator base station data including positioning data of the first user; determining a user in an area represented by the positioning data as the resident member, and determining service subscription information of the resident member in the service system as the second target service subscription information.

6. The method of claim 1, wherein, Before the recommending the target service to the first user, the method further includes: obtaining a first probability distribution of a service type of the target service in the service system and a second probability distribution of a service resource amount of the target service in the service system; calculating a similarity of the first probability distribution and the second probability distribution by a Hayashi distance algorithm to obtain a Hayashi distance value, the Hayashi distance value being used to represent a diversity distribution degree of the target service in the service system; The recommending the target service to the first user includes: in a case where the Hayashi distance value is greater than or equal to a preset distance value, recommending the target service to the first user.

7. A service recommendation apparatus characterized by comprising: includes: The generating module is configured to, in a case where the first service subscription information of the first user is obtained, generate a first service recommendation list based on a first service in the first service subscription information, the first user being a user who subscribes to a service in a service system for the first time, and the first service recommendation list including a second service in the service system having a subscription relationship with the first service. determining a second service recommendation list according to first service subscription information of the first user and second service subscription information of a second user in the service system, the second user including at least one of the following: a family member of the first user, a corresponding resident member of the first user, the resident member including at least one of the following: a daytime resident member, a nighttime resident member, the second service recommendation list including a third service in the second service subscription information; adjusting, by a correction parameter corresponding to the first service, score information of each service in the second service and the third service respectively, to obtain score information of each service; recommending a target service to the first user, the target service being a service in the first service and the second service whose score information satisfies a preset score condition.

8. A computer device, comprising: The device comprises a processor and a memory storing computer program instructions. The processor executes the computer program instructions to implement the service recommendation method of any one of claims 1-6.

9. A readable storage medium, characterized by, The program or instructions are stored on the readable storage medium, and the program or instructions are executed by the processor to implement the steps of the service recommendation method of any one of claims 1-6.

10. A computer program product, characterised in that, The program product is stored in a non-transitory storage medium, and the program product is executed by at least one processor to implement the steps of the service recommendation method of any one of claims 1-6.

Images