CN114840767A

CN114840767A - Service recommendation method based on artificial intelligence and related equipment

Info

Publication number: CN114840767A
Application number: CN202210593914.8A
Authority: CN
Inventors: 刘育基
Original assignee: Ping An Property and Casualty Insurance Company of China Ltd
Current assignee: Ping An Property and Casualty Insurance Company of China Ltd
Priority date: 2022-05-27
Filing date: 2022-05-27
Publication date: 2022-08-02

Abstract

The application provides a service recommendation method and device based on artificial intelligence, an electronic device and a storage medium, wherein the service recommendation method based on artificial intelligence comprises the following steps: analyzing a service scene and a preset service scheme to obtain a rated service index; acquiring user data from a service database corresponding to the service scene based on the rated service index to construct a target user set; classifying each piece of user data in the target user set to obtain a plurality of user clusters; grouping the user data in each user cluster to obtain a user grouping confidence; and evaluating the service scheme corresponding to the service scene based on the user grouping confidence, and recommending the service scheme according to the evaluation result. According to the method, the users are classified for multiple times so as to reduce the characteristic directivity of the users, and therefore the accuracy of service recommendation can be improved.

Description

Service recommendation method based on artificial intelligence and related equipment

Technical Field

The present application relates to the field of artificial intelligence technologies, and in particular, to a service recommendation method and apparatus based on artificial intelligence, an electronic device, and a storage medium.

Background

With the development of the digitization process, an online operation mode is opened in more and more business fields, for example, various mobile phone end application programs with consultation and transaction functions are introduced in the financial industry and the business industry, enterprises can recommend service schemes of updated versions in the application programs to users at irregular intervals to attract the users to use the mobile phone end application programs to increase revenues, and as the number of online users increases day by day, accurate service scheme recommendation to the users becomes more important.

At present, users are generally selected randomly to perform service scheme experiments to determine a calibration service scheme to be recommended, however, the service recommendation method does not consider that the randomly extracted clients have characteristic directivity, and further results of service recommendation are not accurate enough.

Disclosure of Invention

In view of the foregoing, it is desirable to provide a service recommendation method based on artificial intelligence and related devices, so as to solve the technical problem of how to improve the accuracy of service recommendation, where the related devices include an artificial intelligence based service recommendation apparatus, an electronic device and a storage medium.

The embodiment of the application provides a service recommendation method based on artificial intelligence, which comprises the following steps:

analyzing a service scene and a preset service scheme to obtain a rated service index, wherein the service scene refers to a problem to be solved by an enterprise, the service scheme refers to a solution designed by the enterprise aiming at the service scene, and the rated service index is used for representing the quality degree of the service scheme;

acquiring user data from a service database corresponding to the service scene based on the rated service index to construct a target user set;

classifying each piece of user data in the target user set to obtain a plurality of user clusters;

grouping the user data in each user cluster to obtain a user grouping confidence;

and evaluating the service scheme corresponding to the service scene based on the user grouping confidence, and recommending the service scheme according to the evaluation result.

In the service recommendation method, the target user set participating in the test is determined by analyzing the service scene, the target user set is divided into a plurality of user clusters by a clustering method, then the user data in the user clusters are divided into different categories and recombined into a plurality of user groups, and the service recommendation result is evaluated based on the confidence of each group of users, so that the characteristics of each group of users are balanced as much as possible, and the accuracy of the subsequent service recommendation result is improved.

In some embodiments, the collecting user data from a service database corresponding to the service scenario based on the rated service index to construct a target user set includes:

marking user data in a service database corresponding to the service scene according to the service index, and if the service index in the user data is smaller than the rated service index, marking the user data as a target user;

counting the characteristics of the target user, wherein the characteristics comprise user age, user gender, user online time and user preference categories;

and taking the characteristics corresponding to each target user as user data, and storing the user data as the target user set.

Therefore, by comparing the service index of the user data with the rated service index, the target user is collected and marked from the service database and stored as the target user set, so that sample support is provided for subsequent user grouping and service recommendation, and the accuracy of service recommendation is improved.

In some embodiments, the characteristics of the target user include numerical data and non-numerical data, and after the counting the characteristics of the target user, the method further includes:

distinguishing whether each user characteristic belongs to non-numerical data or not, and marking each user characteristic according to the distinguishing result, wherein the marking comprises 'yes' and 'no';

traversing the marks of each user characteristic in sequence, counting the number of the value types of the characteristic if the mark of one user characteristic is 'yes', marking each type in the characteristic in sequence according to a natural number, taking the marked natural number as the code value of each data in the characteristic to obtain a numerical characteristic, and not performing any operation if the mark of one user characteristic is 'no';

each data in the user profile marked "yes" is replaced with its corresponding code value to update the target set of users.

Therefore, the non-numerical data in the characteristics of the user data are coded to obtain fully quantized data, sample support can be provided for subsequent service recommendation, and accuracy of service recommendation can be improved.

In some embodiments, the classifying each piece of user data in the target user set to obtain a plurality of user clusters includes:

calculating the cosine distance between every two pieces of user data according to a cosine distance algorithm;

classifying the target user set according to the cosine distance and a preset clustering algorithm to obtain a plurality of user clusters, wherein each user cluster comprises a plurality of pieces of user data;

respectively counting the number and the polymerization degree of user data in each user cluster, wherein the polymerization degree is used for representing the diversity degree of the characteristics of the user data;

and inputting the quantity and the polymerization degree of the user data into a self-defined integration model to obtain an integration result, and taking the integration result as the weight of the user cluster.

Therefore, a plurality of user clusters are obtained by classifying the user data sets, the number and the polymerization degree of the user data in each user cluster are respectively counted, the weight of each user cluster is obtained according to the number and the polymerization degree, the data can be divided into a plurality of clusters with characteristic similarity, the weight is given to each cluster to represent the importance of the user data in each cluster during service recommendation, and the accuracy of subsequent service recommendation can be improved.

In some embodiments, the custom integration model satisfies the following relationship:

wherein, T _i Representing the weight of the ith user cluster, wherein the higher the weight is, the higher the reliability of the test result is when the service scheme is tested by using the user data in the cluster is indicated; a. the _i Representing the amount of user data in the ith user cluster, said A _i The larger the value of (b) indicates that the more user data is included in the cluster, the higher the weight of the cluster should be; b is _i Represents the aggregation level of the ith user cluster, and the lower the value of the aggregation level, the more discrete the characteristics of the user data in the cluster, the higher the weight of the cluster should be.

Therefore, the integration result of the number and the polymerization degree of the user data is calculated through the user-defined integration model, the value of the integration result can represent the degree that the user data in the user cluster has characteristic directivity, and data support can be provided for the subsequent calculation of the user grouping confidence, so that the accuracy of service recommendation can be improved.

In some embodiments, the grouping the user data in each user cluster to obtain the user grouping confidence comprises:

carrying out secondary classification on the user data in each user cluster by using a preset clustering algorithm, and marking the user data in the user cluster according to the classification, wherein the marking comprises 'experiment' and 'comparison';

respectively calculating the aggregation degree of the user data in each category of each user cluster as an aggregation value;

taking the product of the weight corresponding to each user cluster and the aggregation value as the confidence corresponding to each category;

combining the user data with the same mark to serve as an experimental group and a control group;

and calculating the confidence sum of the user data in the experimental group as the confidence of the experimental group, and calculating the confidence sum of the data in the control group as the confidence of the control group.

Therefore, a plurality of classes of users are obtained by classifying the users in each user cluster, the confidence coefficient of each class is calculated according to the aggregation degree of the user data in each class, and the higher the confidence coefficient value is, the higher the confidence coefficient is, the higher the reliability of the obtained service recommendation result is when the group of users are used as a sample for service recommendation, so that the accuracy of subsequent service recommendation can be improved.

In some embodiments, the evaluating the service scenario corresponding to the service scenario based on the user grouping confidence and recommending the service scenario according to an evaluation result includes:

respectively calculating the average value of the service indexes of the user data in the experimental group and the control group to be used as the reference value of each group of users;

a service scheme is randomly pushed for each group of users, and the mean value of the service indexes of each group of users is respectively counted after a preset test period to serve as a test value;

calculating the difference value between the test value and the reference value to respectively obtain the service increment of each group of users;

calculating the product of the confidence coefficient and the service increment of each group of users to be used as the service recommendation result of each group of users;

and taking the service scheme corresponding to the larger service recommendation result as the recommended calibration service scheme.

Therefore, by pushing apps of different versions to two groups of users and obtaining user online time feedback, the service increment of each group of users is further corrected based on the confidence of each group to obtain a final test result, and the accuracy of a service recommendation result is further improved.

The embodiment of the present application further provides a service recommendation device based on artificial intelligence, the device includes:

the acquisition unit is used for analyzing the service scene and a preset service scheme to acquire a rated service index;

the acquisition unit is used for acquiring user data from a service database corresponding to the service scene based on the rated service index so as to construct a target user set;

a classifying unit, configured to classify each piece of data in the target user set to obtain a plurality of user clusters;

the grouping unit is used for grouping the user data in each user cluster to acquire a user grouping confidence;

and the recommending unit is used for evaluating the service scheme corresponding to the service scene based on the confidence coefficient so as to obtain a service recommending result.

An embodiment of the present application further provides an electronic device, where the device includes:

a memory storing computer readable instructions; and

a processor executing computer readable instructions stored in the memory to implement the artificial intelligence based service recommendation method.

The embodiment of the present application further provides a computer-readable storage medium, in which computer-readable instructions are stored, and the computer-readable instructions are executed by a processor in an electronic device to implement the artificial intelligence based service recommendation method.

In the service recommendation method, the target user set participating in the test is determined by analyzing the service scene, the target user set is divided into a plurality of user clusters by a clustering method, then the user data in the user clusters are divided into different categories and recombined into an experimental group and a comparison group, and the service recommendation result is evaluated and corrected based on the confidence of each group of users, so that the characteristics of each group of users are balanced as much as possible, and the accuracy of subsequent service recommendation is improved.

Drawings

FIG. 1 is a flow chart of a preferred embodiment of an artificial intelligence based service recommendation method to which the present application relates.

Fig. 2 is a functional block diagram of a preferred embodiment of an artificial intelligence based service recommendation apparatus according to the present application.

Fig. 3 is a schematic structural diagram of an electronic device according to a preferred embodiment of the artificial intelligence based service recommendation method according to the present application.

Fig. 4 is a schematic structural diagram of a target user set to which the present application relates.

Detailed Description

For a clearer understanding of the objects, features and advantages of the present application, reference is made to the following detailed description of the present application along with the accompanying drawings and specific examples. It should be noted that the embodiments and features of the embodiments of the present application may be combined with each other without conflict. In the following description, numerous specific details are set forth to provide a thorough understanding of the present application, and the described embodiments are merely a subset of the embodiments of the present application and are not intended to be a complete embodiment.

Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, features defined as "first", "second", may explicitly or implicitly include one or more of the described features. In the description of the present application, "a plurality" means two or more unless specifically limited otherwise.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein in the description of the present application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.

The present application provides an artificial intelligence based service recommendation method, which can be applied to one or more electronic devices, where the electronic device is a device capable of automatically performing numerical calculation and/or information processing according to a preset or stored instruction, and hardware of the electronic device includes, but is not limited to, a microprocessor, an Application Specific Integrated Circuit (ASIC), a programmable gate array (FPGA), a Digital Signal Processor (DSP), an embedded device, and the like.

The electronic device may be any electronic product capable of performing human-computer interaction with a user, for example, a personal computer, a tablet computer, a smart phone, a Personal Digital Assistant (PDA), a game machine, an interactive web television (IPTV), an intelligent wearable device, and the like.

The electronic device may also include a network device and/or a user device. The network device includes, but is not limited to, a single network server, a server group consisting of a plurality of network servers, or a cloud computing (cloud computing) based cloud consisting of a large number of hosts or network servers.

The network where the electronic device is located includes, but is not limited to, the internet, a wide area network, a metropolitan area network, a local area network, a Virtual Private Network (VPN), and the like.

Fig. 1 is a flowchart illustrating a preferred embodiment of the artificial intelligence based service recommendation method according to the present application. The order of the steps in the flow chart may be changed and some steps may be omitted according to different needs.

And S10, analyzing a service scene and a preset service scheme to obtain a rated service index, wherein the service scene refers to a problem to be solved by an enterprise, the service scheme refers to a solution designed by the enterprise aiming at the service scene, and the rated service index is used for representing the quality of the service scheme.

In this optional embodiment, the service scenario may be a problem that needs to be solved by an enterprise during operation, for example, the service scenario in this scheme may refer to a problem that a certain type of APP under the enterprise flag suffers from a decrease in user activity in a past period, where the APP is called Application in its entirety, and its meaning is mobile phone software. The user activity refers to the online time of the user in a certain APP.

In this optional embodiment, the business scheme may be a solution designed by an enterprise for the business scenario, for example, the business scheme in this scheme may refer to modification and update of the page of the APP for the enterprise to improve the user activity of the APP, the preset business scheme may include a scheme R1 and a scheme R2, the scheme R1 may represent a current design scheme of the APP, the scheme R2 may represent another design scheme obtained by modifying the design of the APP, and the design scheme of the APP may refer to a display mode of an image of the page of the user in the APP or a moving mode of a transition animation.

In this alternative embodiment, since the business scenario is that a certain APP under the enterprise flag encounters a problem of decreasing user activity, the nominal business index may be an average online TIME _ total of all users in the past 30 natural days in a history of the APP, where the average online TIME means an average of a sum of total online TIMEs of each user in the past 30 natural days, and the "online TIME" means a TIME span from each TIME when the user turns on the APP to when the APP is turned off. Generally, the longer the online time of a certain type of APP is, the higher the attention or recognition degree of the user to the APP is, the higher the possibility that the enterprise obtains profit through the APP is, and therefore, the online time of each user in the past 30 days can be used as the business index of the user. The service index may be represented by time to represent a total online time in a history of a user, for example, if a certain user logs in the APP five times in the past 30 days, and the online time after each login is 20 seconds, the total online time of the user in the past 30 days is 100 seconds, that is, the service index represents the user as 100 seconds.

In this alternative embodiment, the historical average online TIME of all users in the APP represented by TIME _ total may be used as the rated service index, and the online TIME of each user represented by TIME in the past 30 days may be used as the service index corresponding to each user.

Therefore, a rated service index is obtained by analyzing the service scene and the preset service scheme, the rated service index can represent the concerned quantitative data in the service scene, and can provide data support in subsequent service recommendation so as to improve the accuracy of a service recommendation result.

S11, collecting user data from the service database corresponding to the service scene based on the rated service index to construct a target user set.

In an optional embodiment, the collecting user data from a service database corresponding to the service scenario based on the rated service index to construct a target user set includes:

and S111, marking the user data in the service database corresponding to the service scene according to the service index, and if the service index in the user data is smaller than the rated service index, marking the user data as a target user.

In this optional embodiment, the database is a data set designed, stored, and managed according to a data structure, for example, the service database in this solution may be a memSQL database, which is an open-source database and has a function of storing user data corresponding to the APP.

In this optional embodiment, the specific way of marking the user data in the service database according to the service index is to run a preset first program in the memSQL database to obtain a program return value, where the preset program may be an SQL statement, and may be in the form of a "select 'table' from 'database' where the term < TIME _ total", where 'table' represents a result returned by the preset program, and 'database' represents a database where the target user set is located, a keyword "where" in the SQL statement represents that conditional screening needs to be performed when data is collected, where the condition is "TIME < TIME _ total", where TIME represents a service index of a user to represent a total online TIME in a history of a certain user, and TIME _ total represents the rated service index, and since the screening condition in the preset program is "TIME < TIME _ total", therefore, all the user data in the table represent users whose service indexes are smaller than the rated service index, and therefore, each piece of data in the table can be marked as the target user.

And S112, counting the characteristics of the target user, wherein the characteristics comprise user age, user gender, user online time and user preference categories.

In this optional embodiment, the characteristics of the target user may be counted, and a specific implementation manner of the method may be to run a preset second program in the service database to obtain a program return value, where the form of the second program may be "selectable, b, c, and stable2 free". Wherein a may represent the age characteristic of the target user, b may represent the gender characteristic of the target user, c may represent the online duration of the target user, i.e., the service index corresponding to each target user, d may represent the preference category of the target user, and table2 may represent a data table composed of the characteristics of the target user.

In this optional embodiment, the characteristics of the target user include numerical data and non-numerical data, and after the statistics of the characteristics of the target user, the method further includes:

In this alternative embodiment, whether each user feature belongs to non-numerical data may be identified according to a preset Python program, taking the age feature represented by a as an example, the preset Python program may be in the form of "print (a ═ int)", which means that a data format of the age feature represented by a is compared with an "integer" data format, and a comparison result is output, where the comparison result includes "YES" and "NO", and if the comparison result is "YES", the age feature represented by a is marked as "NO", and if the comparison result is "NO", the age feature represented by a is marked as "YES".

In this optional embodiment, the features of the user data may be sequentially traversed to obtain the tag thereof, and if the tag of a certain column of features is "yes", the number of value categories of the data in the feature is counted, each category is sequentially tagged according to a natural number, and the natural number tag is used as a code corresponding to the data in the feature.

For example, taking the sex characteristics represented by b as an example, the sex characteristics include { woman, man }, and then the natural numbers are used to mark "woman" as 1 and mark "man" as 2.

In this alternative embodiment, the natural number flag corresponding to each piece of data in the non-numerical type feature may be used as the encoded value corresponding to the piece of data, and each piece of data may be replaced with its corresponding encoded value.

And S113, taking the characteristics corresponding to each target user as user data, and storing the user data as the target user set.

In this alternative embodiment, the features corresponding to each target user may be arranged column by column and a target user set may be constructed, where the target user set may be in the form of a data table, where each row represents a piece of user data, and where each column represents a feature of the user data.

Illustratively, the first column in the target set of users may represent the age of the user, a, which is a positive integer and the range of values for a may be [0, + ∞ ]; the second column may represent the gender b of the user, with b having a value of {1,2 }; the third column may represent the user's online time period c within the APP, c being a positive integer and c may range from 0, + ∞; the fourth column may represent the user's preferred content d, which may range in value [1, + ∞ ]. Illustratively, fig. 4 is a schematic structural diagram of the target user set.

Therefore, by marking the characteristics of each user data counted by the target user and storing the characteristics as a target user set, the non-numerical data in the characteristics of the user data are coded to obtain fully quantized data, sample support can be provided for subsequent service recommendation, and the accuracy of service recommendation can be improved.

S12, classifying each piece of user data in the target user set to obtain a plurality of user clusters.

In an optional embodiment, the classifying each piece of user data in the target user set to obtain a plurality of user clusters includes:

and S121, calculating the cosine distance between every two pieces of user data according to a cosine distance algorithm.

In this alternative embodiment, each piece of user data may be regarded as a code vector according to the arrangement order of the features of the user data, for example, the arrangement order of the features of the user data may be a, b, c, d, and then the code vector of one user data may be [20,2,360,1 ].

In this alternative embodiment, the distance between each user datum may be calculated according to a preset distance metric algorithm, where the preset distance metric algorithm may be a cosine distance algorithm, for example, if a vector a corresponding to a certain user datum is ═ a _A ,b _A ,c _A ,d _A ]And the other user data corresponds to a vector B ═ a _B ,b _B ,c _B ,d _B ]Then, the specific calculation method of the cosine distance between the vectors a and B is:

wherein d is _A,B Represents the cosine similarity between vector a and vector B, with part a · B representing the inner product of vector a and vector B, | a | representing the modulo length of vector a, and | B | representing the modulo length of vector B.

The calculation mode of the inner product and the module length is as follows:

A·B＝a _A ×a _B +b _A ×b _B +c _A ×c _B +d _A ×d _b

where subscripts a and B represent vectors of feature membership for a dimension.

Illustratively, when a ═ 1,2,3,4]And B ═ 2,3,4,5]When d is above _A,B The calculation method is as follows:

the cosine distance between the vector a and the vector B is 0.01.

In this alternative embodiment, the cosine distance between the user data is used to represent the similarity between two users, and if the distance between two user data is larger, it indicates that the similarity between the two users is lower.

And S122, classifying the target user set according to the cosine distance and a preset clustering algorithm to obtain a plurality of user clusters, wherein each user cluster comprises a plurality of pieces of user data.

In this optional embodiment, the target user set may be classified according to a preset clustering algorithm to obtain a plurality of user clusters, where the specific implementation flow of the preset clustering algorithm is as follows:

a1: marking all user data in the target user set as unvisited;

a2: randomly selecting one user data marked as unvisited and marking as X, and further marking the user data represented by the X as the visited;

a3: selecting the subsequent steps according to a preset judging condition, wherein the preset judging condition is that the cosine distance between the user data represented by the X and the user data is smaller than or equal to a preset radius d _e Has n pieces of user data, if n is less than a preset threshold minpts, proceed to step A4, otherwise proceed to step A12, exemplarily, said d _e May be 0.5, said minpts may be 4;

a4: creating a new cluster, recording the new cluster as C, and adding the X user into the C;

a5: recording the user data represented by the X as a center and the radius as d _e Is N;

a6: randomly selecting one piece of user data in the N and recording the user data as Y;

a7: if the label of Y is unisited, marking Y as visited;

a8: if the radius is d by taking the user data represented by the Y as the center _e If there are at least minpts objects in the neighborhood, adding these objects to the set N;

a9: adding said Y to said cluster C if said Y does not already belong to any cluster;

a10: repeating the steps A6 to A10 until N is an empty set;

a11: marking the cluster C as a first user cluster and marking the cluster C as a cluster C ⁱ Where i represents the number of cycles and the initial value of i may be set to 1;

a12: marking X as a noise point;

a13: repeating the steps A2-A13 until all points in the original user data set are marked as visited and the algorithm ends and obtains a plurality of user clusters, and recording the set of the plurality of user clusters as C _z And can be recorded as C _z ＝[C ¹ ,C ² ,…,C ⁿ ]Wherein n is said C _z The number of clusters in the cluster.

And S123, respectively counting the number and the polymerization degree of the user data in each user cluster, wherein the polymerization degree is used for representing the diversity degree of the characteristics of the user data.

In this alternative embodiment, the number of user data in each user cluster may be counted, and for example, the number of user data in each cluster may be denoted as a ═ a ¹ ,A ² ,…,A ⁿ And b, wherein a represents a set of the number of the user data, each element in a corresponds to the number of the user data in a certain user cluster, and a superscript of each element represents an index of the user cluster corresponding to the element.

In this optional embodiment, the variance of the user data in each user cluster may be calculated to serve as the aggregation degree of the user data in each user cluster, where a larger variance indicates a larger difference between the user data in the user cluster, and a lower aggregation degree indicates that the user data in the user cluster has no obvious characteristic directivity, so that a test result obtained when the user data in the user cluster is used for service recommendation is more accurate.

And S124, inputting the number and the polymerization degree of the user data into a self-defined integration model to obtain an integration result, and using the integration result as the weight of the user cluster.

In this optional embodiment, the amount and the polymerization degree of the user data may be input into a customized integration model to obtain an integration result, where the customized integration model satisfies the following relation:

wherein, T _i Representing the integration result of the number and the polymerization degree of the user data in the ith user cluster, wherein the higher the value of the integration result is, the higher the reliability of the test result is when the user data in the cluster is used as a sample to test the service scheme; a. the _i Representing the amount of user data in the ith user cluster, said A _i A larger value of (d) indicates that more user data is included in the cluster, the higher the importance of the cluster should be; b is _i Represents the aggregation level of the ith user cluster, and the lower the value of the aggregation level, the more discrete the characteristics of the user data in the cluster are, the higher the importance of the cluster should be.

Illustratively, when A _i ＝20，B _i When the value is 5, the T is _i The representative integrated results were calculated as:

in this alternative embodiment, the integration result T of the number of user data and the aggregation degree in each user cluster may be used as the weight of the corresponding user cluster.

And S13, grouping the user data in each user cluster to obtain the user grouping confidence.

In an optional embodiment, the grouping the user data in each user cluster to obtain the user grouping confidence includes:

s131, performing secondary classification on the user data in each user cluster by using a preset clustering algorithm, and marking the user data in the user cluster according to the categories, wherein the marking comprises 'experiment' and 'comparison'.

In this alternative embodiment, each user cluster may be divided into two categories to obtain two categories of user data. The preset clustering algorithm can be a K-means clustering algorithm, and the specific implementation process comprises the following steps:

a1, setting a classification number K according to the number of the service schemes, for example, K is 2 in the scheme;

a2, randomly selecting K user data from all user data of a user cluster as centroids, wherein, for example, the K centroids can be respectively denoted as K1 and K2 in the present scheme;

a3, traversing each user data in the user cluster, and calculating the euclidean distance between each user data and each centroid respectively, illustratively, the euclidean distance between a certain user data and K1 is D1, and the euclidean distance between the certain user data and K2 is D2, if D1 is smaller than D2, the user data is classified into the category where K1 is located, otherwise, the user data is classified into the category where K2 is located, and after the traversal is finished, two categories can be obtained, each category respectively containing a plurality of user data;

a4, calculating the mean value of the user data in each category and recording as K1_ mean and K2_ mean, calculating the Euclidean distance between the user data in each category and the mean value in the corresponding category, and selecting the user data with the minimum Euclidean distance as a new centroid corresponding to each category and recording as K1_ new and K2_ new;

a5, calculating the Euclidean distance between K1_ new and K1 and recording the Euclidean distance as D1_ new, calculating the Euclidean distance between K2_ new and K2 and recording the Euclidean distance as D2_ new, and recording a preset distance threshold as Dt;

a6, if D1_ new and D2_ new are both smaller than Dt, the algorithm is terminated, the category where K1 is located and the category where K2 is located are used as final classification results, otherwise, the steps A3 to A6 are repeatedly executed until the algorithm is terminated.

In this alternative embodiment, the classification result of each user cluster may be obtained based on the execution flow of the Kmeans algorithm, each user cluster may be divided into two classes, and each class in each user cluster may be denoted as G ═ G ^{1 experiment} ,G ^{1 control} ),(G ^{2 experiment} ,G ^{2 control} ),…,(G ^{n experiment} ,G ^{n control} )]Exemplary, G ^{1 experiment} Representing the category labeled "experiment" in the user cluster with index 1, G ¹ ^Control Representing categories marked as "controls" in the user cluster with index 1, where each category contains multiple user data.

S132, respectively calculating the aggregation degree of the user data in each category of each user cluster as an aggregation value.

In this alternative embodiment, the variance of the user data of each category in each user cluster may be calculated, where the variance is used to characterize the aggregation level of the users in each category, and the calculation formula of the variance is:

the S represents the variance of the user data in a certain category and is used for representing the aggregation degree of the user data in the category, the higher the value of the S is, the lower the aggregation degree of the user data in the category is, and the user data in the category has stronger characteristic directivity; u represents the amount of user data in the category; j represents an index of the user data in the category; x generationThe value of one user data in the category is shown;

representing the mean of the user data in that category.

Illustratively, when u is 3, x ₁ ＝[1,2,3,4]，x ₂ ＝[2,3,4,5]，x ₃ ＝[3,4,5,6]Then, the calculation mode of S is as follows:

in this optional embodiment, the reciprocal of the variance S may be used as an aggregation value corresponding to each category in each user cluster, and a higher aggregation value indicates that the characteristic directivity of the user data in the corresponding category is lower, so that the more accurate the test result obtained when the user data in the category is used as a sample for service recommendation is.

And S133, taking the product of the weight corresponding to each user cluster and the aggregation value as the confidence corresponding to each category.

In this alternative embodiment, the product of the weight corresponding to each user cluster and the aggregation value may be used as the confidence corresponding to each category, where the confidence satisfies the following relation:

wherein, t ^we Representing the confidence degree corresponding to the category marked as e in the user cluster with index w, wherein the higher the confidence degree value is, the higher the confidence degree of the user data in the category is, S ^we Representing the variance, T, corresponding to the user data in that category ^w Representing the confidence of the cluster of users with index w.

Exemplary, when S ¹ Fruit of Chinese wolfberry ^{Test (experiment)} ＝3.622，T ¹ When t is 0.982, the ^we The representative confidence is calculated as follows:

in this alternative embodiment, the t may be ^we As the confidence corresponding to each class, the t ^we The higher the value of (A) is, the higher the reliability of the obtained test result is when the user data in the category is used as a sample for service recommendation.

S134, the user data with the same label are combined to be used as an experimental group and a control group.

In this alternative embodiment, the categories having the same label in all the user clusters may be combined into an experimental group and a control group, and for example, the categories may be represented as:

experimental group ═ G ^{1 experiment} ,G ^{2 experiment} ,…,G ^{n experiment} ]；

Control group ═ G ^{1 control} ,G ^{2 control} ,…,G ^{n control} ]。

In this alternative embodiment, both the experimental group and the control group contain a plurality of user data.

And S135, calculating the confidence sum of the user data in the experimental group as the confidence of the experimental group, and calculating the confidence sum of the data in the control group as the confidence of the control group.

In this optional embodiment, the confidence sum of the user data in the experimental group may be calculated as the confidence of the experimental group, for example, if there are 3 user clusters in the solution and t is t ^{1 experiment} ＝0.271，t ^{2 experiment} ＝0.28，t ^{3 experiment} When the confidence score is 0.29, the confidence score of the experimental group is calculated as follows:

t experiment 0.271+0.28+0.29 ═ 0.841

In this alternative embodiment, the confidence sum of the user data in the comparison group may be calculated as the confidence of the comparison group, for example, if there are 3 user clusters in the present solution and t is t ^{1 control} ＝0.29，t ^{2 control} ＝0.289，t ^{3 control} If 0.35, the confidence of the control group is calculated as:

t control 0.29+0.289+ 0.35-0.929

In this alternative embodiment, the normalized confidence levels of the experimental group and the control group may be calculated according to a maximum normalization algorithm, and for example, when the T experiment is 0.841 and the T control is 0.929, the normalized confidence levels of the experimental group may be calculated as follows:

the normalized confidence of the control group is calculated in the following manner:

in this alternative embodiment, the normalized confidence levels may be used as the user grouping confidence levels for the experimental group and the control group, respectively.

And S14, evaluating the service scheme corresponding to the service scene based on the user grouping confidence, and recommending the service scheme according to the evaluation result.

In this optional embodiment, the evaluating the service scheme corresponding to the service scenario based on the user grouping confidence and recommending the service scheme according to the evaluation result includes:

and S141, respectively calculating the average value of the service indexes of the user data in the experimental group and the control group to be used as the reference value of each group of users.

In this optional embodiment, the average of the business indexes of the user data in the experimental group and the control group may be calculated respectively, where the business indexes areFor example, when the experimental group includes 10 users and the average of the online durations in the user data in the experimental group is 300 seconds, the online duration in the user data may be recorded as the online duration in the user data

When the control group comprises 20 users and the average value of the online time length in the user data in the control group is 360 seconds, the time length can be recorded as

In this alternative embodiment, the

As a reference value for the users in the experimental group, the value of (c) can be used

As a baseline value for the users in the control group.

And S142, randomly pushing a service scheme for each group of users, and respectively counting the average value of the service indexes of each group of users after a preset test period to serve as a test value.

In this optional embodiment, the enterprise may push apps with the R1 business scheme to the users in the experimental group, push apps with the R2 business scheme to the users in the control group, and count the average online duration of the users in the experimental group after a preset test period, and record the average online duration as

And counting the average online time of the users in the control group and recording the average online time

In this alternative embodiment, the

Value of asThe test values of the experimental groups are described

The value of (d) was used as the test value of the control group.

And S143, calculating the difference value between the test value and the reference value to respectively obtain the service increment of each group of users.

In this alternative embodiment, the business increment of the experimental group may be recorded as V ^{Experiment of} The business increment of the comparison group can be recorded as V ^Control The calculation mode of the service increment of the experimental group is as follows:

in this optional embodiment, the calculation method of the service increment of the comparison group is as follows:

in this alternative embodiment, the V may be ^{Experiment of} As the business increment of the experimental group, the value of V may be set ^Control As the traffic increment of the control group.

And S144, calculating the product of the confidence coefficient and the service increment of each group of users to serve as the service recommendation result of each group of users.

In this alternative embodiment, the confidence level Tg may be based ^{Experiment of} And Tg ^Control Correcting the service increment of each user group to obtain a service recommendation result of each group of users, wherein the calculation method of the service recommendation result comprises the following steps:

R ^{experiment of} ＝Tg ^{Experiment of} ×V ^{Experiment of}

R ^Control ＝Tg ^Control ×V ^Control

Wherein, Tg ^{Experiment of} Confidence, Tg, representing the experimental group ^Control Representing the confidence of the control groupDegree, V ^{Experiment of} Traffic increment, V, representing the experimental group ^Control Traffic increment, R, representing said control group ^{Experiment of} Representing the result of the business recommendation, R, of the experimental group ^Control Representing a service recommendation result of the control group.

And S145, taking the service scheme corresponding to the larger service recommendation result as the recommended calibration service scheme.

In this alternative embodiment, R may be compared to ^{Experiment of} Value of (D) and the value of R ^{Control of} The business scheme corresponding to the larger value is used as a calibration business scheme for enterprise popularization.

Illustratively, when R is ^{Experiment of} 10 and R ^Control If 20, the business scheme pushed by the users in the contrast group by the enterprise is the R2 business scheme, the R2 business scheme may be used as a calibration business scheme for enterprise promotion.

Fig. 2 is a functional block diagram of a preferred embodiment of an artificial intelligence based service recommendation apparatus according to an embodiment of the present application. The artificial intelligence based service recommendation device 11 comprises an acquisition unit 110, a collection unit 111, a classification unit 112, a grouping unit 113 and a recommendation unit 114. The module/unit referred to in this application refers to a series of computer program segments that can be executed by the processor 13 and that can perform a fixed function, and that are stored in the memory 12. In the present embodiment, the functions of the modules/units will be described in detail in the following embodiments.

In an optional embodiment, the obtaining unit 110 is configured to analyze a service scenario and a preset service scheme to obtain a rated service index, where the service scenario refers to a problem that needs to be solved by an enterprise, the service scheme refers to a solution that is designed by the enterprise for the service scenario, and the rated service index is used to represent a degree of superiority and inferiority of the service scheme.

In this optional embodiment, the business scheme may be a solution designed by an enterprise for the business scenario, for example, the business scheme in the scheme may refer to modification and update of the page of the APP by the enterprise in order to improve the user activity of the APP, the preset business scheme may include a scheme R1 and a scheme R2, the scheme R1 may represent a current design scheme of the APP, the scheme R2 may represent another design scheme obtained by modifying the design of the APP, and the design scheme of the APP may refer to a display mode of an image of the user page in the APP or a moving mode of a scene cut animation.

In an optional embodiment, the collecting unit 111 is configured to collect user data from a service database corresponding to the service scenario based on the rated service index to construct a target user set.

Illustratively, the first column in the target set of users may represent the age of the user, a, which is a positive integer and the range of values for a may be [0, + ∞ ]; the second column may represent the gender b of the user, the value of b being {1,2 }; the third column may represent the user's online time period c within the APP, c being a positive integer and c may range from 0, + ∞; the fourth column may represent the user's preferred content d, which may range in value [1, + ∞ ]. Illustratively, fig. 4 is a schematic diagram of the target user set.

In an alternative embodiment, the classifying unit 112 is configured to classify each piece of user data in the target user set to obtain a plurality of user clusters.

In this optional embodiment, the classifying each piece of user data in the target user set to obtain a plurality of user clusters includes:

and inputting the quantity and the polymerization degree of the user data into a self-defined integration model to obtain an integration result, and using the integration result as the weight of the user cluster.

In this alternative embodiment, the distance between each user data may be calculated according to a preset distance metric algorithm, which may be a cosine distance algorithm, for example,if the vector A corresponding to a certain user data is ═ a _A ,b _A ,c _A ,d _A ]And the other user data corresponds to a vector B ═ a _B ,b _B ,c _B ,d _B ]Then, the specific calculation method of the cosine distance between the vectors a and B is as follows:

The calculation mode of the inner product and the module length is as follows:

A·B＝a _A ×a _B +b _A ×b _B +c _A ×c _B +d _A ×d _b

the cosine distance between the vector a and the vector B is 0.01.

a1: marking all user data in the target user set as unvisited;

a4: creating a new cluster and marking as C, and adding the X user into C;

a7: if the label of the Y is unisited, marking the Y as visited;

a10: repeating the steps A6 to A10 until N is an empty set;

a11: recording the cluster C as a first user cluster and marking the cluster C as a cluster C ⁱ Where i represents the number of cycles and the initial value of i may be set to 1;

a12: marking X as a noise point;

a13: repeating said steps A2 toA13, until all points in the original user data set are marked as visited, the algorithm is ended and a plurality of user clusters are obtained, and the set of the user clusters is marked as C _z And can be recorded as C _z ＝[C ¹ ,C ² ,…,C ⁿ ]Wherein n is said C _z The number of clusters in the cluster.

In this optional embodiment, the number and the degree of polymerization of the user data may be input into a customized integration model to obtain an integration result, where the customized integration model satisfies the following relation:

wherein, T _i Representing the integration result of the number and the polymerization degree of the user data in the ith user cluster, wherein the higher the value of the integration result is, the higher the reliability of the test result is when the user data in the cluster is used as a sample to test the service scheme; a. the _i Representing the amount of user data in the ith user cluster, said A _i A larger value of (a) indicates that more user data is included in the cluster, the higher the importance of the cluster should be;B _i represents the aggregation level of the ith user cluster, and the lower the value of the aggregation level, the more discrete the characteristics of the user data in the cluster are, the higher the importance of the cluster should be.

Exemplary, when A _i ＝20，B _i When the value is 5, the T is _i The representative integrated results were calculated as:

in this alternative embodiment, the integration result T of the number and aggregation degree of the user data in each user cluster may be used as the weight of the corresponding user cluster.

In an alternative embodiment, the grouping unit 113 is configured to group the user data in each user cluster to obtain a user grouping confidence.

In this optional embodiment, the grouping the user data in each user cluster to obtain the user grouping confidence includes:

combining the user data with the same label to serve as an experimental group and a control group;

In this alternative embodiment, the classification result of each user cluster may be obtained based on the execution flow of the Kmeans algorithm, each user cluster may be divided into two categories, and for each category in each user cluster, the category may be denoted as G ═ G [ (G) ^{1 experiment} ,G ^{1 control} ),(G ^{2 experiment} ,G ^{2 control} ),…,(G ^{n experiment} ,G ^{n control} )]Exemplary, G ^{1 experiment} Representing the category labeled "experiment" in the user cluster with index 1, G ¹ ^Control Representative index is1, wherein each category contains a plurality of user data.

the S represents the variance of the user data in a certain category and is used for representing the aggregation degree of the user data in the category, the higher the value of the S is, the lower the aggregation degree of the user data in the category is, and the user data in the category has stronger characteristic directivity; u represents the amount of user data in the category; j represents an index of the user data in the category; x represents the value of one of the user data in the category;

representing the mean of the user data in that category.

Exemplary, when S ^{1 experiment} ＝3.622，T ¹ When t is 0.982, the ^we The representative confidence is calculated as follows:

Control group ═ G ^{1 control} ,G ^{2 control} ,…,G ^{n control} ]。

In this alternative embodiment, the confidence sum of the user data in the experimental group may be calculated as the confidence of the experimental group, for example, if there are 3 user clusters in the scheme and t is t ^{1 experiment} ＝0.271，t ^{2 experiment} ＝0.28，t ^{3 experiment} When the confidence score is 0.29, the confidence score of the experimental group is calculated as follows:

t experiment 0.271+0.28+0.29 ═ 0.841

t control 0.29+0.289+ 0.35-0.929

In an optional embodiment, the recommending unit 114 is configured to evaluate a service scenario corresponding to the service scenario based on the user grouping confidence, and recommend the service scenario according to an evaluation result.

In this optional embodiment, evaluating the service scenario corresponding to the service scenario based on the user grouping confidence, and recommending the service scenario according to the evaluation result includes:

randomly pushing a service scheme for each group of users, and respectively counting the mean value of the service indexes of each group of users after a preset test period to be used as a test value;

In this optional embodiment, the average of the service indicators of the user data in the experimental group and the control group may be calculated respectively, where the service indicator is the online time length in the user data, and for example, when the experimental group includes 10 users and the average of the online time lengths in the user data in the experimental group is 300 seconds, the average may be recorded as the online time length in the user data in the experimental group

In this alternative embodiment, the

As a baseline value for the users in the control group.

In this optional embodiment, the enterprise may push apps with the R1 business scheme to users in the experimental group, push apps with the R2 business scheme to users in the control group, and count the average online duration of the users in the experimental group after a preset test period and record the average online duration as

In this alternative embodiment, the

As a test value of the experimental group, the value of

The value of (d) was used as the test value of the control group.

in this alternative embodiment, the V may be ^{Experiment of} The value of (a) is used as the traffic increment of the experimental group, and the value of the Vcontrol can be used as the traffic increment of the control group.

R ^{experiment of} ＝Tg ^{Experiment of the invention} ×V ^{Experiment of}

R ^Control ＝Tg ^Control ×V ^Control

Wherein, Tg ^{Experiment of} Confidence, Tg, representing the experimental group ^Control Representing the confidence of the control group, V ^{Experiment of} Traffic increment, V, representing the experimental group ^Control Representing the traffic increment, R, of said control group ^{Experiment of} Representing the result of the business recommendation, R, of the experimental group ^Control Representing a service recommendation result of the control group.

In this alternative embodiment, R may be compared ^{Experiment of} Value of (D) and the value of R ^Control The business scheme corresponding to the larger value is used as a calibration business scheme for enterprise popularization.

Fig. 3 is a schematic structural diagram of an electronic device according to an embodiment of the present application. The electronic device 1 comprises a memory 12 and a processor 13. The memory 12 is used for storing computer readable instructions, and the processor 13 is used for executing the computer readable instructions stored in the memory to implement the artificial intelligence based service recommendation method according to any one of the above embodiments.

In an alternative embodiment, the electronic device 1 further comprises a bus, a computer program stored in the memory 12 and executable on the processor 13, such as an artificial intelligence based service recommendation program.

Fig. 3 only shows the electronic device 1 with components 12-13, and it will be understood by a person skilled in the art that the structure shown in fig. 3 does not constitute a limitation of the electronic device 1, and may comprise fewer or more components than shown, or a combination of certain components, or a different arrangement of components.

In conjunction with fig. 1, the memory 12 in the electronic device 1 stores a plurality of computer-readable instructions to implement an artificial intelligence based service recommendation method, and the processor 13 can execute the plurality of instructions to implement:

Specifically, the specific implementation method of the instruction by the processor 13 may refer to the description of the relevant steps in the embodiment corresponding to fig. 1, which is not described herein again.

It will be understood by those skilled in the art that the schematic diagram is merely an example of the electronic device 1, and does not constitute a limitation to the electronic device 1, the electronic device 1 may have a bus-type structure or a star-type structure, and the electronic device 1 may further include more or less hardware or software than those shown in the figures, or different component arrangements, for example, the electronic device 1 may further include an input and output device, a network access device, etc.

It should be noted that the electronic device 1 is only an example, and other existing or future electronic products, such as those that may be adapted to the present application, should also be included in the scope of protection of the present application, and are included by reference.

Memory 12 includes at least one type of readable storage medium, which may be non-volatile or volatile. The readable storage medium includes flash memory, removable hard disks, multimedia cards, card type memory (e.g., SD or DX memory, etc.), magnetic memory, magnetic disks, optical disks, etc. The memory 12 may in some embodiments be an internal storage unit of the electronic device 1, for example a removable hard disk of the electronic device 1. The memory 12 may also be an external storage device of the electronic device 1 in other embodiments, such as a plug-in mobile hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a flash memory card (FlashCard), and the like, provided on the electronic device 1. Further, the memory 12 may also include both an internal storage unit and an external storage device of the electronic device 1. The memory 12 may be used not only to store application software installed in the electronic device 1 and various types of data, such as codes of artificial intelligence-based service recommendation programs, etc., but also to temporarily store data that has been output or is to be output.

The processor 13 may be composed of an integrated circuit in some embodiments, for example, a single packaged integrated circuit, or may be composed of a plurality of integrated circuits packaged with the same or different functions, including one or more Central Processing Units (CPUs), microprocessors, digital processing chips, graphics processors, and combinations of various control chips. The processor 13 is a control core (control unit) of the electronic device 1, connects various components of the whole electronic device 1 by using various interfaces and lines, and executes various functions of the electronic device 1 and processes data by running or executing programs or modules (e.g., executing an artificial intelligence based service recommendation program, etc.) stored in the memory 12 and calling data stored in the memory 12.

The processor 13 executes an operating system of the electronic device 1 and various types of application programs installed. The processor 13 executes the application program to implement the steps in each of the artificial intelligence based service recommendation method embodiments described above, such as the steps shown in fig. 1.

Illustratively, the computer program may be partitioned into one or more modules/units that are stored in the memory 12 and executed by the processor 13 to accomplish the present application. The one or more modules/units may be a series of computer-readable instruction segments capable of performing certain functions, which are used to describe the execution of the computer program in the electronic device 1. For example, the computer program may be divided into an acquisition unit 110, an acquisition unit 111, a classification unit 112, a grouping unit 113, a recommendation unit 114.

The integrated unit implemented in the form of a software functional module may be stored in a computer-readable storage medium. The software functional module is stored in a storage medium and includes several instructions to enable a computer device (which may be a personal computer, a computer device, or a network device) or a processor (processor) to execute parts of the artificial intelligence based service recommendation method according to the embodiments of the present application.

The integrated modules/units of the electronic device 1 may be stored in a computer-readable storage medium if they are implemented in the form of software functional units and sold or used as separate products. Based on such understanding, all or part of the processes in the methods of the embodiments described above may be implemented by a computer program, which may be stored in a computer-readable storage medium and executed by a processor, to implement the steps of the embodiments of the methods described above.

Wherein the computer program comprises computer program code, which may be in the form of source code, object code, an executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, usb disk, removable hard disk, magnetic disk, optical disk, computer memory, Read-only memory (ROM), random access memory and other memory, etc.

Further, the computer-readable storage medium may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function, and the like; the storage data area may store data created according to the use of the blockchain node, and the like.

The block chain referred by the application is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.

The bus may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one arrow is shown in FIG. 3, but this does not indicate only one bus or one type of bus. The bus is arranged to enable connected communication between the memory 12 and at least one processor 13 or the like.

Although not shown, the electronic device 1 may further include a power supply (such as a battery) for supplying power to each component, and preferably, the power supply may be logically connected to the at least one processor 13 through a power management device, so that functions such as charge management, discharge management, and power consumption management are implemented through the power management device. The power supply may also include any component of one or more dc or ac power sources, recharging devices, power failure detection circuitry, power converters or inverters, power status indicators, and the like. The electronic device 1 may further include various sensors, a bluetooth module, a Wi-Fi module, and the like, which are not described herein again.

Further, the electronic device 1 may further include a network interface, and optionally, the network interface may include a wired interface and/or a wireless interface (such as a WI-FI interface, a bluetooth interface, etc.), which are generally used for establishing a communication connection between the electronic device 1 and other electronic devices.

Optionally, the electronic device 1 may further comprise a user interface, which may be a Display (Display), an input unit (such as a Keyboard), and optionally a standard wired interface, a wireless interface. Alternatively, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (organic light-emitting diode) touch device, or the like. The display, which may also be referred to as a display screen or display unit, is suitable for displaying information processed in the electronic device 1 and for displaying a visualized user interface, among other things.

The embodiment of the present application further provides a computer-readable storage medium (not shown), in which computer-readable instructions are stored, and the computer-readable instructions are executed by a processor in an electronic device to implement the artificial intelligence based service recommendation method according to any of the above embodiments.

It is to be understood that the described embodiments are for purposes of illustration only and that the scope of the appended claims is not limited to such structures.

In the several embodiments provided in the present application, it should be understood that the disclosed system, apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the modules is only one logical functional division, and other divisions may be realized in practice.

The modules described as separate parts may or may not be physically separate, and parts displayed as modules may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment.

In addition, functional modules in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, or in a form of hardware plus a software functional module.

Furthermore, it will be obvious that the term "comprising" does not exclude other elements or steps, and the singular does not exclude the plural. A plurality of units or means recited in the specification may also be implemented by one unit or means through software or hardware. The terms first, second, etc. are used to denote names, but not any particular order.

Finally, it should be noted that the above embodiments are only used for illustrating the technical solutions of the present application and not for limiting, and although the present application is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions can be made on the technical solutions of the present application without departing from the spirit and scope of the technical solutions of the present application.

Claims

1. A service recommendation method based on artificial intelligence is characterized by comprising the following steps:

2. The artificial intelligence based service recommendation method of claim 1, wherein the collecting user data from the service database corresponding to the service scenario based on the rated service index to construct a target user set comprises:

3. The artificial intelligence based service recommendation method of claim 2, wherein the characteristics of the target user comprise numerical data and non-numerical data, and after counting the characteristics of the target user, the method further comprises:

4. The artificial intelligence based service recommendation method of claim 1, wherein said classifying each piece of user data in said target user set to obtain a plurality of user clusters comprises:

5. The artificial intelligence based service recommendation method of claim 4, wherein said customized integration model satisfies the following relation:

6. The artificial intelligence based service recommendation method of claim 1, wherein said grouping user data in each user cluster to obtain a user grouping confidence comprises:

7. The artificial intelligence based service recommendation method according to claim 6, wherein said evaluating the service scenario corresponding to the service scenario based on the user grouping confidence and recommending the service scenario according to the evaluation result comprises:

8. An artificial intelligence based service recommendation apparatus, the apparatus comprising:

and the recommending unit is used for evaluating the service scheme corresponding to the service scene based on the user grouping confidence and recommending the service scheme according to the evaluation result.

9. An electronic device, characterized in that the electronic device comprises:

a memory storing computer readable instructions; and

a processor executing computer readable instructions stored in the memory to implement the artificial intelligence based service recommendation method of any of claims 1-7.

10. A computer-readable storage medium characterized by: the computer readable storage medium stores computer readable instructions which are executed by a processor in an electronic device to implement the artificial intelligence based service recommendation method according to any one of claims 1 to 7.