CN114493644A

CN114493644A - Method and system for recommending 5G terminal equipment and sales information thereof to user

Info

Publication number: CN114493644A
Application number: CN202011268410.6A
Authority: CN
Inventors: 刘健; 魏丫丫; 罗仕漳; 仲籽彦; 张健; 汪利伟; 张明哲
Original assignee: China Telecom Corp Ltd
Current assignee: China Telecom Corp Ltd
Priority date: 2020-11-13
Filing date: 2020-11-13
Publication date: 2022-05-13

Abstract

The disclosure relates to a method for recommending 5G terminal equipment and sales information thereof to a user, which comprises the following steps: the 5G terminal equipment predicts the preference of the user based on the historical data of the user and the related users; acquiring information of a hall store which is close to the user and sells the 5G terminal equipment preferred by the user; and pushing the information of the 5G terminal equipment preferred by the user and the information of the hall shop selling the 5G terminal equipment preferred by the user to the user. The present disclosure also relates to a system and a computer system for recommending 5G terminal devices and sales information thereof to a user, and a computer-readable storage medium.

Description

Method and system for recommending 5G terminal equipment and sales information thereof to user

Technical Field

The disclosure relates to a method and a system for recommending 5G terminal equipment and sales information thereof to a user.

Background

The 5G network has three characteristics of high speed, large connection and low time delay, and the preference users of the 5G terminal also have obvious difference. The existing 5G terminal marketing mainly depends on an online channel or offline channel single-channel sales mode on one hand, and on the other hand, the differentiation scene requirements of high speed and high stability of a 5G mobile terminal user are not considered, although the effect can be improved through sales promotion activities, the problems that online and offline channels are separated and split, target users are not accurate, and marketing effect is limited still exist. Therefore, how to realize accurate marketing based on user preference and online-offline fusion achieves the goals of cost reduction and efficiency improvement, and is a key research direction of operators.

Disclosure of Invention

The following presents a simplified summary of the disclosure in order to provide a basic understanding of some aspects of the disclosure. However, it should be understood that this summary is not an exhaustive overview of the disclosure. It is not intended to identify key or critical elements of the disclosure or to delineate the scope of the disclosure. Its sole purpose is to present some concepts of the disclosure in a simplified form as a prelude to the more detailed description that is presented later.

According to a first aspect of the present disclosure, there is provided a method for recommending 5G terminal equipment and sales information thereof to a user, comprising: the 5G terminal equipment predicts the preference of the user based on the historical data of the user and the related users; acquiring information of a hall store which is close to the user and sells the 5G terminal equipment preferred by the user; and pushing the information of the 5G terminal equipment preferred by the user and the information of the hall shop selling the 5G terminal equipment preferred by the user to the user.

According to a second aspect of the present disclosure, there is provided a system for recommending 5G terminal devices and sales information thereof to a user, comprising: the prediction module is configured to predict the 5G terminal equipment preferred by the user based on historical data of the user and related users; a hall store information acquisition module configured to acquire information of hall stores in the vicinity of the user that sell the 5G terminal device preferred by the user; and the information pushing module is configured to push the information of the 5G terminal equipment preferred by the user and the information of the hall shop selling the 5G terminal equipment preferred by the user to the user.

According to a third aspect of the present disclosure, there is provided a computer system recommending 5G terminal devices and sales information thereof to a user, comprising: a memory having instructions stored thereon; and a processor configured to execute instructions stored on the memory to perform a method according to the above aspects of the disclosure.

According to a fourth aspect of the present disclosure, there is provided a computer-readable storage medium comprising computer-executable instructions that, when executed by one or more processors, cause the one or more processors to perform a method according to the above-mentioned aspects of the present disclosure.

The method is based on a 5G differentiated application scene, combines operator stock user data to accurately predict 5G terminal preference, and then guides a user to experience and finish sales in a nearby 5G terminal sales hall store, so that online and offline fused marketing is realized, and the aim of marketing effect is improved.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the disclosure and together with the description, serve to explain the principles of the disclosure.

The present disclosure may be more clearly understood from the following detailed description with reference to the accompanying drawings, in which:

fig. 1 is a flowchart of a method for recommending 5G terminal devices and sales information thereof to a user according to an embodiment of the present disclosure.

Fig. 2 is a block diagram of a system for recommending 5G terminal devices and sales information thereof to a user according to an embodiment of the present disclosure.

Fig. 3 is a schematic diagram of prediction in a method of recommending 5G terminal devices and sales information thereof to a user according to an embodiment of the present disclosure.

Fig. 4 is a structural diagram of a prediction module in a system for recommending 5G terminal devices and sales information thereof to a user according to an embodiment of the present disclosure.

Fig. 5 is a schematic diagram of obtaining store information in a hall in a method for recommending 5G terminal equipment and sales information thereof to a user according to an embodiment of the present disclosure.

Fig. 6 is a flowchart of a method for recommending 5G terminal devices and sales information thereof to a user according to an embodiment of the present disclosure.

Fig. 7 is a block diagram of an exemplary configuration of a computing device capable of implementing embodiments in accordance with the present disclosure.

Detailed Description

The following detailed description is made with reference to the accompanying drawings and is provided to assist in a comprehensive understanding of various exemplary embodiments of the disclosure. The following description includes various details to aid understanding, but these details are to be regarded as examples only and are not intended to limit the disclosure, which is defined by the appended claims and their equivalents. The words and phrases used in the following description are used only to provide a clear and consistent understanding of the disclosure. In addition, descriptions of well-known structures, functions, and configurations may be omitted for clarity and conciseness. Those of ordinary skill in the art will recognize that various changes and modifications of the examples described herein can be made without departing from the spirit and scope of the disclosure.

Fig. 1 is a flowchart of a method 100 for recommending a 5G terminal device and sales information thereof to a user according to an embodiment of the present disclosure. Fig. 6 is a flowchart of a method for recommending 5G terminal equipment and sales information thereof to a user according to a specific embodiment of the present disclosure. The method for recommending 5G terminal equipment and sales information thereof to a user according to the present disclosure is described below with reference to fig. 1 and 6.

The method 100 includes the steps of: 5G terminal equipment for predicting user preference based on historical data of the user and related users (step S110); acquiring information of a hall store of a 5G terminal device in the vicinity of the user, which sells user preferences (step S120); and pushing information of the 5G terminal device preferred by the user and information of the hall shop selling the 5G terminal device preferred by the user to the user (step S130). The various steps of method 100 are described in detail below.

Step 110: and 5G terminal equipment for predicting user preference based on historical data of the user and related users.

Fig. 3 is a schematic diagram of a 5G terminal device for predicting the user preference based on the historical data of the user and the related users. And acquiring historical data of the user and related users for prediction from different data sources, performing fusion analysis on the data, and predicting the preference of the user through a pre-established prediction model. The predictive model is built based on historical data of a large number of users. For example, a user who has changed the terminal of 5G may be extracted from the history data as a positive sample, and model training may be performed based on the following six categories of information.

The historical data for the user and associated users used to make the prediction may include: user basic information, user internet surfing information, user attribute information, user terminal information, user preference information, user package information and the like. Such information may be obtained from different information sources. As shown in fig. 3, for example, the user basic information and the user internet access information may be obtained from a Customer Relationship Management (CRM) system, the user attribute information and the user preference information may be obtained from a pre-stored client behavior database, the user terminal information may be obtained from a terminal management platform, and the user package information may be obtained from a billing system of an operator.

Each of the categories of information described above may include one or more metrics. For example, the user base information may include, but is not limited to: mobile phone number, standard province ID, local network ID, title, age, value group, product instance state, main sale type, customer star level, etc. The user internet information may include, but is not limited to: the most preferred mobile terminal Application (APP), the second preferred APP, the third preferred APP, the fourth preferred APP, the fifth preferred APP, whether traffic is overflowing users, etc. The user attribute information may include, but is not limited to: family grouping, professional grouping, social grouping, traffic saturation, associated credit ratings (e.g., whether there is a debt, amount of communication service consumption, etc.), whether there is a vehicle group, whether there is a group abroad, etc. The user terminal information may include, but is not limited to: historical number of changes, number of changes in recent 12 months, average duration of last 2 years, average duration of last, duration of last of one terminal of the user (duration of last refers to the time from purchase to replacement of the terminal, which indicates the market in which the user owns the terminal), duration of last of another terminal of the user, frequency preference of changes, number of brands owned, etc. The user preference information may include, but is not limited to: a new-line preference, a holiday-exchange preference, a preferred terminal price type (e.g., low-end type, luxury type, noodle type, frugal type, etc.), a preferred terminal brand type (e.g., whether to prefer an apple phone, whether to prefer a chinese phone, whether to prefer a millet phone, etc.). User package information may include, but is not limited to: average value of ARPU (average communication service income contributed by each user) in the last 6 months, average value of ARPU in the last 3 months, average value of total number of calls in the last 3 months, average value of total time-consuming duration in the last 3 months, average value of number of calls in the last 3 months, average value of charging duration of calls in the last 3 months, average value of package flow in the last 3 months, and the like.

As described above, the predictive model may be created by extracting the users who have changed the machine 5G terminal from the history data as positive samples (and also extracting the users who have not changed the machine 5G terminal as negative samples), and performing model training based on the above-mentioned six categories of information of these users. Considering the complexity of 5G terminal prediction, the prediction model may include three parts, namely a first model for predicting whether the potential 5G terminal user is a 5G terminal exchange user, a second model for predicting the price of the 5G terminal preferred by the user, and a third model for predicting the brand of the 5G terminal preferred by the user. In one embodiment, the above-described first to third models, i.e., a first classification model for predicting whether the user has a willingness to change the phone, a second classification model for predicting the price of the 5G terminal preferred by the user, and a third classification model for predicting the brand of the 5G terminal preferred by the user, may be implemented with classification models. The mathematical model on which the models are based can be selected according to the data condition and the training effect. In one embodiment, the first classification model is trained based on a random forest model, and the second and third classification models are trained based on corresponding multi-layered perceptron models. The first to third classification models are described in detail below, respectively.

First classification model

A first classification model for predicting whether a user has a willingness to change the machine (i.e., a willingness to purchase a 5G terminal) is trained based on a random forest model. In model training, a sample set for training a model may be established. Each sample in the sample set comprises an input and an output, the six categories of information (wherein the information of each category comprises one or more indexes) input for the respective sample user, and the output is whether the user purchases a 5G terminal. In step 110, that is, when the model is used to predict whether the current user intends to change the machine, the input of the first classification model is the above six categories of information of the predicted user (where each category of information includes one or more indicators), and the output is whether the current user intends to change the machine. For example, an output of 0 represents that the user is predicted to have no intention to change the machine, and an output of 1 represents that the user is predicted to have an intention to change the machine.

The process of training the model by using the input and output samples comprises the following steps: adjusting model parameters of the random forest model (for example, traversing all values of each parameter in a parameter adjusting range and traversing combinations of the values of each parameter), taking recall rate, accuracy and F1 value as model evaluation standards, and finally selecting F1 value as an evaluation index to determine the optimal parameters of the random forest model (namely selecting the model parameters capable of enabling the F1 value to be the maximum as the optimal parameters) to complete training of the random forest model, thereby establishing the first classification model.

The parameter adjusting range of the random forest model comprises the following steps:

the maximum depth max _ depth of the tree is in the range of [6, 12);

the learning rate learning _ rate is [0.05,0.2], and values are taken every 0.05;

the value of the tree n _ estimators is [50,200], and the value is taken every 10.

The finally determined optimal parameters of the random forest model are as follows:

{'n_estimators':170,

'max_depth':8,

'learning_rate':0.05}。

some samples may be collected to build a test sample set that evaluates the performance of the model. For example, a proportion (e.g., 1/5) of the samples from the sample set may be extracted to create a test sample set. The evaluation can be made by three evaluation criteria, i.e., recall, accuracy, and F1 value. Wherein the meaning of each evaluation criterion is as follows:

the number of people who predicted to purchase the airplane and actually purchased the airplane/the number of people who actually purchased the airplane

The accuracy rate is the number of people with correct prediction/the number of people with correct prediction plus wrong prediction

F1 value 2 recall accuracy/(recall + accuracy)

The values of the respective evaluation criteria for testing the first classification model with the above-determined optimal parameters are as follows:

recall rate	49.1％
		Rate of accuracy	81.3％
F1 value	61.2％

Second classification model

The second classification model for predicting the price of the 5G terminal preferred by the user is trained based on the multi-layer perceptron model. In model training, a sample set for training a model may be established. Each sample in the sample set includes an input and an output, the six categories of information input for the respective sample user (where the information for each category includes one or more metrics), and the price output for the 5G terminal purchased for that user. In step 110, when the model is used to predict the price of the 5G terminal preferred by the current user, the input of the second classification model is the above six categories of information (where the information of each category includes one or more indexes) of the predicted user, which outputs the price of the 5G terminal preferred by the current user.

To simplify the model and its input and output, the price may be divided into intervals. For example, the price of a 5G terminal is divided by 1000 and rounded down as the output of the model. This means that in the set of samples created, the output (i.e. price) of each sample is the price of the 5G terminal purchased by the user divided by 1000 and rounded down; and, at the time of prediction, the output of the second classification model is the price of the 5G terminal preferred by its predicted current user divided by 1000 and rounded down. For example, the output of the second classification model is 0, which represents that the predicted price interval of the 5G terminal preferred by the current user is 0-999 yuan, the output of the second classification model is 1, which represents that the predicted price interval of the 5G terminal preferred by the current user is 1000-1999 yuan, the output of the second classification model is 2, which represents that the predicted price interval of the 5G terminal preferred by the current user is 2000-2999 yuan, and so on. Each value output by the second classification model is also referred to as a class hereinafter.

The process of training the model by using the input and output samples comprises the following steps: adjusting model parameters of the multilayer perceptron model (for example, traversing all values of the parameters in the following parameter adjustment range and traversing combinations of the values of the parameters), taking a macro recall rate, a macro accuracy rate and a macro F1 value as model evaluation criteria, and finally selecting a macro F1 value as an evaluation index to determine the optimal parameters of the multilayer perceptron model (namely selecting the model parameters capable of enabling the macro F1 value to be the maximum as the optimal parameters) to complete training of the multilayer perceptron model, thereby establishing a second classification model.

The parameter adjusting range of the multilayer perceptron model comprises the following steps:

the hidden _ layer _ sizes are in the range of [50,100], and values are taken every 10;

solver under { 'lbfgs', 'sgd', 'adam' };

activation:{‘logistic’,‘tanh’,‘relu’}；

max _ iter is [200,1000], taking values every 50;

alpha is taken at [1e-5,0.01] every 0.001.

The finally determined optimal parameters of the multilayer perceptron model are as follows:

MLPClassifier(hidden_layer_sizes＝(100,50),solver＝'lbfgs',activation＝‘logistic’,alpha＝1e-5,random_state＝123,max_iter＝1000)。

some samples may be collected to build a test sample set that evaluates the performance of the model. For example, a proportion (e.g., 1/5) of the samples from the sample set may be extracted to create a test sample set. The evaluation can be made by three evaluation criteria, namely, macro recall, macro accuracy, and macro F1 values. Wherein the meaning of each evaluation criterion is as follows:

macroscopic recall-the average of recall of all categories

Macroscopic accuracy-the average of the accuracies of all classes

Macroscopic F1 value 2 macroscopic recall ratio macroscopic accuracy/(macroscopic recall ratio + macroscopic accuracy)

Wherein the meaning of the recall and accuracy of each of the categories can be referred to the above description of the evaluation criteria of the first classification model.

The values of the respective evaluation criteria for testing the second classification model with the above-determined optimal parameters are as follows:

macroscopic recall	70％
		Macroscopic accuracy	54％
Macroscopic viewF1 value	57％

Third classification model

The third classification model of the brand of the 5G terminal for predicting the user preference is trained based on a multi-layer perceptron model. In model training, a sample set for training a model may be established. Each sample in the sample set includes an input and an output, the six categories of information input for the respective sample user (where the information for each category includes one or more metrics), and the brand of the 5G terminal purchased for that user. In performing step 110, that is, when the model is used to predict the brand of the 5G terminal preferred by the current user, the input of the third classification model is the above six categories of information (where the information of each category includes one or more indexes) of the predicted user, which is output as the brand of the 5G terminal preferred by the current user. For example, the output of the third classification model is 0 for brand 1, 1 for brand 2, 2 for brand 3, and so on. Each value output by the third classification model is also referred to as a class hereinafter.

The process of training the model by using the input and output samples comprises the following steps: adjusting model parameters of the multilayer perceptron model (for example, traversing all values of the parameters in the following parameter adjustment range and traversing combinations of the values of the parameters), taking a macro recall rate, a macro accuracy rate and a macro F1 value as model evaluation criteria, and finally selecting a macro F1 value as an evaluation index to determine the optimal parameters of the multilayer perceptron model (namely selecting the model parameters capable of enabling the macro F1 value to be the maximum as the optimal parameters) to complete training of the multilayer perceptron model, thereby establishing a third classification model.

solver under { 'lbfgs', 'sgd', 'adam' };

activation:{‘logistic’,‘tanh’,‘relu’}；

max _ iter is [200,1000], taking values every 50;

alpha is taken at [1e-5,0.01] every 0.001.

macroscopic recall-the average of recall of all categories

Macroscopic accuracy-the average of the accuracies of all classes

The values of the respective evaluation criteria for testing the third classification model with the above-determined optimum parameters are as follows:

macroscopic recall	52％
		Macroscopic accuracy	45％
Macroscopic F1 value	45％

Prediction model

And the prediction model scores each 5G terminal model to be recommended according to the results output by the first to third classification models, and obtains the predicted 5G terminal equipment preferred by the user according to the scores.

In response to the output of the first classification model indicating that the user has a willingness to change, calculating a price score f-score of the 5G terminal model to be recommended according to the following formula_Price：

Wherein the predicted price is indicated by an output result of the second classification model, and a brand score f-score of the 5G terminal model to be recommended is calculated according to the following formula_Brand：

Wherein the predicted brand is indicated by an output of the third classification model. Wherein the first push brand refers to a brand that is most desirable to recommend according to the current marketing objective, and the second push brand refers to a brand that is next desirable to recommend according to the current marketing objective. In one particular example, the first push owner brand may include Huanwan and HONOR, and the second push owner brand may include millet, VIVO, and OPPO.

Then, a composite score f-score of the 5G terminal model to be recommended is calculated according to the following formula_{5G terminal model}：

f-score_{5G terminal model}＝f-score_Price+f-score_BrandSo as to complete the scoring of the 5G terminal model to be recommended. Finally, the composite score f-score is calculated_{5G terminal model}And determining the highest 5G terminal model to be recommended as the predicted user preferred 5G terminal equipment.

In an actual recommendation operation, the predicted results of a plurality of users are usually obtained in batch. In one particular example, step 110 is performed periodically (e.g., weekly) to make the prediction. For example, each time the information of each user updated in the period time can be input into the first to third classification models, the prediction model will obtain the list of the predicted 5G terminal switch users and the 5G terminal information of their preferences. An example of a particular prediction result is shown in the following table:

serial number	5G terminal model	Target user list
			1	Hua is P40 pro	216 ten thousand target audience
2	Huachen nova 7pro	161 ten thousand target audience
			3	Millet 10pro	183 ten thousand target audience
4	Huashi Mate30	248 ten thousand target audience
			5	vivo Z6	91 ten thousand target audience
6	oppo A92s	77 ten thousand target audience

Step 120: information of a hall store in the vicinity of the user selling the 5G terminal device preferred by the user is acquired.

The method comprises the steps that information of a hall store in a preset range nearby a user is obtained based on geographic position data of the user and longitude and latitude data of a 5G terminal sales hall store; inquiring whether 5G terminal equipment preferred by a user is sold or not from each hall store; and sorting the hall stores selling the 5G terminal equipment preferred by the user according to the distance to the user, and determining the nearest hall store as the hall store to be recommended.

Fig. 5 illustrates the principle of this step. A specific example is described below in conjunction with fig. 5.

First, the predicted geographic Location of the 5G terminal switch user (also referred to herein as a potential switch user, or potential 5G end user) obtained in step 110 is obtained (e.g., the user's geographic Location is obtained via LBS (Location Based Services) Location data of the user's terminals), which is denoted as (X1, Y1). Then, the information management system of the sales hall stores of the 5G terminal acquires latitude and longitude data of each hall store, and records the latitude and longitude data as (X2, Y2).

The distance D from the user to the store is calculated according to the following formula:

d ═ R × arcos [ cos (Y1) × (Y2) × (X1-X2) + sin (Y1) × sin (Y2) ], where R is the earth radius and R ═ 6371 KM.

For example, according to the geographic position of the user Y, the preset range is set to be 5KM, and the following three hall store information are screened out according to a distance formula:

china telecom apricot stone mouth business hall (Xishan Yingfu business center): 1.8 KM;

china telecom business office (North Master & university culture school district): 2.25 KM; and

china telecom wing Internet mobile phone store (Yiyuan creative base): 2.9 KM.

Then, based on the information of the 5G terminal of the user preference predicted in step 110, the terminal is queried in the above-described hall stores. For example, the 5G terminal preference information of the user Y is "hua is MATE 30", and "hua is MATE 30" is inquired one by one in the three hall stores (i.e., the chinese telecommunications apricot stone opening business hall (west mountain win house business center), the chinese telecommunications communication business hall (north teachers and universities), and the chinese telecommunications sky wing internet mobile phone store (yi yuan creative base)), and the commodity is inquired in the chinese telecommunications apricot stone opening business hall (west mountain win house business center) and the chinese telecommunications sky wing internet mobile phone store (yi yuan creative base) for sale, but the commodity cannot be inquired in the chinese telecommunications apricot stone opening business hall (west mountain win house business center).

And finally, marking the Chinese telecommunication business hall (northern teachers and university culture school zone) with the shorter distance as a hall store to be recommended according to the distance. Exemplary specific results are shown in the following table:

in the step, based on the LBS geographic position data of the user terminal and the longitude and latitude data of the 5G terminal sales hall store, information of the hall store in a preset range near the user is obtained, and the hall stores are sorted according to the distance. Based on the information of the preferred 5G terminal of the user predicted in step 110, if the preferred 5G terminal of the user can be queried from the store within the preset range, the store information is marked as the store information to be recommended.

Step 130: will be provided withInformation of user-preferred 5G terminal device and hall shop selling user-preferred 5G terminal device The information is pushed to the user.

The information pushed to the user of the store shop selling the user preferred 5G terminal devices may include: price information of a user's preferred 5G terminal device sold by a hall store, coupon information (including coupon information, discount information, etc.), address information, route information of the user to the hall store, and contact information of the hall store.

In a specific example, the 5G terminal sales information of the hall store to be recommended, such as the chinese telecom business office (north teachers and universities) obtained in step 120, may be obtained first. For example, the detail information of the Huaqi MATE30 terminal preferred by the user Y is obtained, and the sales price of the shop is 3999 yuan for the Huaqi MATE30 terminal to carry out the direct descending 300 yuan sales promotion. The store terminal sales manager information and contact information, such as the king manager and the phone xxxx thereof, can be further queried. Then, based on the user geographical location information and the hall store geographical location information, calculating route information of the user to the hall store, such as: the inquiry shows that the user Y can arrive at the China telecom business office (North Master and university culture school district) by riding for 11 minutes.

Then, the acquired price information, coupon information, customer manager contact information, route information and the like are integrated into a designed file template, and the information is sent to a mobile phone of a user through an online channel. The online channel may include, for example, short messages, WeChat, APP, and the like.

Optionally, the method 100 may further include: the adoption of the information of the pushed 5G terminal device preferred by the user and the information of the hall store selling the 5G terminal device preferred by the user is taken as the history data of the user for the following prediction (step S140). Through the 5G terminal transaction system, the transaction data of the hall store recommended by the target user in step 130 in the above steps can be acquired, including but not limited to: whether to go to a store, whether to purchase the recommended 5G terminal, and the price, brand, model, transaction completion time, etc. at which the 5G terminal was purchased. The transaction data is recorded as historical data of the user, and the models can be subjected to incremental training and testing as samples so as to continuously optimize the parameters of the models.

Fig. 2 is a block diagram of a system 200 for recommending 5G terminal devices and sales information thereof to a user according to an embodiment of the present disclosure. The system 200 includes a prediction module 210, a store information acquisition module 220, an information push module 230, and a data collection module 240. Wherein, the prediction module 210 is configured to perform the operation of the step 110, the store information acquisition module 220 is configured to perform the operation of the step 120, the information push module 230 is configured to perform the operation of the step 130, and the data collection module 240 is configured to perform the operation of the step 140. Fig. 4 is a block diagram of a prediction module 210 in a system recommending 5G terminal devices and sales information thereof to users according to an embodiment of the present disclosure. The prediction module 210 includes classification models 211, 212, 213 (corresponding to the first to third classification models described above, respectively), and a prediction model 214 (corresponding to the prediction model described above).

A specific example is illustrated below. The prediction module 210 obtains information such as basic information, customer attributes, package information, terminal information, internet access information, user preferences and the like of a user in a preset time period from various data sources, focuses on network preference requirements of high speed and high stability, and predicts the 5G terminal preferences of the user through the classification models 211, 212, 213 and the prediction model 214. The hall information acquisition module 220 acquires hall information within a preset range near the user based on the user terminal LBS geographic position data and the latitude and longitude data of the 5G terminal sales hall store, and sorts the hall stores according to distance; if the 5G terminal preferred by the user can inquire from the hall stores within the preset range, marking the hall store information as the hall store information to be recommended. The information pushing module 230 acquires price information, coupon information, route information and contact information of the 5G terminal preferred by the user from the information of the to-be-recommended hall and stores and sends the price information, the coupon information, the route information and the contact information to the mobile phone of the user. The data collection module 240 obtains 5G terminal transaction information and coupon reimbursement information of a user at a hall store, including hall stores, prices, brands, models, time of purchasing a 5G terminal, and obtains purchasing data (such as purchasing business hall, purchasing time, purchasing brands, prices, and models) of the user for adjusting model parameters in the classification models 211, 212, 213 and the prediction model 214 to optimize system performance.

The 5G network has the characteristics of ultrahigh speed, super-large connection and ultralow time delay, and compared with a 4G user, the user requirements of the 5G mobile terminal have obvious difference, such as the requirements of the user on ultrahigh-definition videos, games, VR/AR and the like. The method and the system are based on the operator stock user data, focus on the network demand preference of the user with high speed and high stability, and innovatively provide an accurate marketing model aiming at the characteristics of the 5G terminal. The high-quality marketing service recommendation method based on the multi-source data of the operator stock users is based on the compliance analysis and the feature extraction (including basic information, customer attributes, package information, terminal information, internet surfing information, user preference and the like) of the multi-source data, and achieves high-quality marketing service recommendation through the integration of model prediction of online data and online store-to-store experience and sales.

Fig. 7 is a block diagram of an exemplary configuration of a computing device 700 capable of implementing embodiments in accordance with the present disclosure. Computing device 700 is an example of a hardware device to which the above-described aspects of the disclosure can be applied. Computing device 700 may be any machine configured to perform processing and/or computing. The computing device 700 may be, but is not limited to, a workstation, a server, a desktop computer, a laptop computer, a tablet computer, a Personal Data Assistant (PDA), a smart phone, an in-vehicle computer, or a combination thereof.

As shown in fig. 7, computing device 700 may include one or more elements that may be connected to or in communication with a bus 702 via one or more interfaces. The bus 2102 may include, but is not limited to, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MCA) bus, an enhanced ISA (eisa) bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus, among others. Computing device 700 may include, for example, one or more processors 704, one or more input devices 706, and one or more output devices 708. The one or more processors 704 may be any kind of processor and may include, but are not limited to, one or more general-purpose processors or special-purpose processors (such as special-purpose processing chips). Input device 706 may be any type of input device capable of inputting information to a computing device and may include, but is not limited to, a mouse, a keyboard, a touch screen, a microphone, and/or a remote controller. Output device 708 may be any type of device capable of presenting information and may include, but is not limited to, a display, speakers, a video/audio output terminal, a vibrator, and/or a printer.

The computing device 700 may also include or be connected to a non-transitory storage device 714, which non-transitory storage device 714 may be any non-transitory and data storage enabled storage device, and may include, but is not limited to, disk drives, optical storage devices, solid state memory, floppy disks, flexible disks, hard disks, magnetic tape, or any other magnetic medium, compact disks or any other optical medium, cache memory, and/or any other memory chip or module, and/or any other medium from which a computer can read data, instructions, and/or code. Computing device 700 may also include Random Access Memory (RAM)710 and Read Only Memory (ROM) 712. The ROM 712 may store programs, utilities or processes to be executed in a nonvolatile manner. The RAM 710 may provide volatile data storage and store instructions related to the operation of the computing device 700. The computing device 700 may also include a network/bus interface 716 that couples to a data link 718. The network/bus interface 716 may be any kind of device or system capable of enabling communication with external devices and/or networks, and may include, but is not limited to, a modem, a network card, an infrared communication device, a wireless communication device, and/or a chipset (such as bluetooth)^TMDevices, 802.11 devices, WiFi devices, WiMax devices, cellular communications facilities, etc.).

The present disclosure may be implemented as any combination of devices, systems, integrated circuits, and computer programs on non-transitory computer readable media. One or more processors may be implemented as an Integrated Circuit (IC), an Application Specific Integrated Circuit (ASIC), or a large scale integrated circuit (LSI), a system LSI, or a super LSI, or as an ultra LSI package that performs some or all of the functions described in this disclosure.

The present disclosure includes the use of software, applications, computer programs or algorithms. Software, applications, computer programs, or algorithms may be stored on a non-transitory computer readable medium to cause a computer, such as one or more processors, to perform the steps described above and depicted in the figures. For example, one or more memories store software or algorithms in executable instructions and one or more processors may associate a set of instructions to execute the software or algorithms to provide various functionality in accordance with embodiments described in this disclosure.

Software and computer programs (which may also be referred to as programs, software applications, components, or code) include machine instructions for a programmable processor, and may be implemented in a high-level procedural, object-oriented, functional, logical, or assembly or machine language. The term "computer-readable medium" refers to any computer program product, apparatus or device, such as magnetic disks, optical disks, solid state storage devices, memories, and Programmable Logic Devices (PLDs), used to provide machine instructions or data to a programmable data processor, including a computer-readable medium that receives machine instructions as a computer-readable signal.

By way of example, computer-readable media can comprise Dynamic Random Access Memory (DRAM), Random Access Memory (RAM), Read Only Memory (ROM), electrically erasable read only memory (EEPROM), compact disk read only memory (CD-ROM) or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to carry or store desired computer-readable program code in the form of instructions or data structures and which can be accessed by a general-purpose or special-purpose computer or a general-purpose or special-purpose processor. Disk or disc, as used herein, includes Compact Disc (CD), laser disc, optical disc, Digital Versatile Disc (DVD), floppy disk and blu-ray disc where disks usually reproduce data magnetically, while discs reproduce data optically with lasers. Combinations of the above are also included within the scope of computer-readable media.

The subject matter of the present disclosure is provided as examples of apparatus, systems, methods, and programs for performing the features described in the present disclosure. However, other features or variations are contemplated in addition to the features described above. It is contemplated that the implementation of the components and functions of the present disclosure may be accomplished with any emerging technology that may replace the technology of any of the implementations described above.

Additionally, the above description provides examples, and does not limit the scope, applicability, or configuration set forth in the claims. Changes may be made in the function and arrangement of elements discussed without departing from the spirit and scope of the disclosure. Various embodiments may omit, substitute, or add various procedures or components as appropriate. For example, features described with respect to certain embodiments may be combined in other embodiments.

In addition, in the description of the present disclosure, the terms "first," "second," "third," and the like are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or order.

Similarly, while operations are depicted in the drawings in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. In some cases, multitasking and parallel processing may be advantageous.

Claims

1. A method for recommending 5G terminal equipment and sales information thereof to a user comprises the following steps:

the 5G terminal equipment predicts the preference of the user based on the historical data of the user and the related users;

acquiring information of a hall store which is close to the user and sells the 5G terminal equipment preferred by the user; and

and pushing the information of the 5G terminal equipment preferred by the user and the information of the hall shop selling the 5G terminal equipment preferred by the user to the user.

2. The method of claim 1, further comprising:

and taking the adoption condition of the user on the pushed information of the 5G terminal equipment preferred by the user and the information of the hall shop selling the 5G terminal equipment preferred by the user as historical data of the user for later prediction.

3. The method of claim 1, wherein the predicting comprises:

predicting a first result including whether the user has a willingness to change a machine based on historical data of the user and related users and a first classification model;

predicting a second result of the price of the 5G terminal including the user preference based on the historical data of the user and the related users and a second classification model;

predicting a third result of the brand of the 5G terminal including the user preference based on historical data of the user and related users and a third classification model; and

and scoring each 5G terminal model to be recommended according to the first to third results, and obtaining the predicted 5G terminal equipment preferred by the user according to the scoring.

4. A method as claimed in claim 3, wherein the first classification model is trained on a random forest model and the second and third classification models are trained on respective multi-layered perceptron models.

5. The method of claim 3, wherein scoring each 5G terminal model to be recommended according to the first through third results comprises: in response to the first result indicating that the user has a willingness to change machines,

calculating a price score f-score of the 5G terminal model to be recommended according to the following formula_Price：

Wherein the predicted price is indicated by the second outcome;

calculating the 5 to be recommended according to the following formulaBrand score f-score of G terminal model_Brand：

Wherein the predicted brand is indicated by the third result; and

calculating the comprehensive score f-score of the 5G terminal model to be recommended according to the following formula_{5G terminal model}：

f-score_{5G terminal model}＝f-score_Price+f-score_BrandAnd finishing the grading of the 5G terminal model to be recommended.

6. The method of claim 5, wherein the 5G terminal device that derives the predicted user preference from the score comprises: (ii) applying the composite score f-score_{5G terminal model}And determining the highest 5G terminal model to be recommended as the predicted 5G terminal equipment preferred by the user.

7. The method of claim 1, wherein the historical data of the user and related users comprises: user basic information, user internet access information, user attribute information, user terminal information, user preference information, and user package information.

8. The method of claim 1, wherein obtaining information of a store near the user that sells the user-preferred 5G terminal device comprises:

acquiring information of a hall shop within a preset range near the user based on the geographic position data of the user and the longitude and latitude data of the 5G terminal sales hall shop;

inquiring whether 5G terminal equipment preferred by the user is sold or not from each hall store; and

and sorting the hall stores selling the 5G terminal equipment preferred by the user according to the distance to the user, and determining the nearest hall store as the hall store to be recommended.

9. The method of claim 1, wherein the information of the hall store selling the user preferred 5G terminal device comprises: price information, preference information, address information, route information of the user to the hall store, and contact information of the hall store of the 5G terminal device preferred by the user sold by the hall store.

10. A system for recommending 5G terminal equipment and sales information thereof to a user comprises:

the prediction module is configured to predict the 5G terminal equipment preferred by the user based on historical data of the user and related users;

a hall store information acquisition module configured to acquire information of hall stores in the vicinity of the user that sell the 5G terminal device preferred by the user; and

and the information pushing module is configured to push the information of the 5G terminal equipment preferred by the user and the information of the hall shop selling the 5G terminal equipment preferred by the user to the user.

11. The system of claim 10, further comprising:

a data collection module configured to take the user's adoption of the pushed information of the user-preferred 5G terminal device and the information of the store selling the user-preferred 5G terminal device as the historical data of the user for the later prediction.

12. The system of claim 10, wherein the prediction module further comprises first through third classification models, wherein the prediction module is further configured to:

predicting a first result including whether the user has a willingness to change based on historical data of the user and related users and a first classification model;

13. The system of claim 12, wherein the first classification model is trained based on a random forest model, and the second and third classification models are trained based on respective multi-layered perceptron models.

14. A computer system for recommending 5G terminal equipment and sales information thereof to a user, comprising:

a memory having instructions stored thereon; and

a processor configured to execute instructions stored on the memory to perform the method of any of claims 1 to 9.

15. A computer-readable storage medium comprising computer-executable instructions that, when executed by one or more processors, cause the one or more processors to perform the method of any one of claims 1-9.