Terminal recommendation method, device, electronic device and storage medium

Abstract

The embodiments of the present invention provide a terminal recommendation method, device, electronic device and storage medium. The terminal recommendation method includes: obtaining user behavior information, base station information and terminal information on sale; generating permanent location information based on the user behavior information; associating the permanent location information with the base station information to determine the target coverage user information; inputting the target coverage user information and the terminal information on sale into a preset terminal prediction model to generate terminal recommendation information; and sending the terminal recommendation information. The user's permanent location is determined through user behavior information, and combined with base station information, users covered by a specific mobile communication network base station are identified; based on the terminal prediction model, the user's terminal preference is predicted and recommended, which significantly reduces the user's selection cost and improves the accuracy of terminal recommendations.

Description

Technical Field

[0001] The present invention relates to the field of big data technology, and in particular to a terminal recommendation method, a terminal recommendation device, an electronic device and a storage medium. Background Art

[0002] In the field of mobile communication technology, many terminal equipment applications rely on the latitude and longitude information of base stations, such as personnel positioning applications, surrounding interest point query applications, etc. Therefore, how to effectively combine the latitude and longitude information of base stations with the information of mobile communication network terminals such as 5G (5th Generation Mobile Communication Technology) and quickly and accurately implement precise marketing recommendations for users and preferences is an urgent problem that major operators need to solve. However, with the advent of 5G mobile communications, the diversification of 5G terminals has brought considerable challenges to user selection, the cost of user selection has increased, and the accuracy of terminal recommendations is low. Summary of the invention

[0003] In view of the above problems, embodiments of the present invention are proposed to provide a terminal recommendation method, a corresponding terminal recommendation device, an electronic device and a storage medium that overcome the above problems or at least partially solve the above problems.

[0004] In a first aspect of the present invention, an embodiment of the present invention discloses a terminal recommendation method, including:

[0005] Obtain user behavior information, base station information, and terminal information on sale;

[0006] generating permanent location information based on the user behavior information;

[0007] Associating the permanent location information with the base station information to determine target coverage user information;

[0008] Inputting the target coverage user information and the on-sale terminal information into a preset terminal prediction model to generate terminal recommendation information;

[0009] The terminal recommendation information is sent.

[0010] Optionally, the method further comprises:

[0011] Obtain user transaction information;

[0012] The model parameters of the terminal prediction model are adjusted based on the user transaction information.

[0013] Optionally, the user behavior information includes location information and traffic usage information, and the step of generating permanent location information based on the user behavior information includes:

[0014] For any location information, determine the longitude and latitude data corresponding to the location information;

[0015] Classify the traffic usage information to generate traffic usage on weekdays and traffic usage on weekends;

[0016] Associating the working day traffic usage with the longitude and latitude data to generate working day location data, and associating the rest day traffic usage with the longitude and latitude data to generate rest day location data;

[0017] The working day location data are sorted according to the size of the working day traffic usage to generate working day residence data; the rest day location data are sorted according to the size of the rest day traffic usage to generate rest day residence data;

[0018] The working daily location data and the resting daily location data are determined as permanent location information.

[0019] Optionally, the step of associating the resident location information with the base station information to determine target coverage user information includes:

[0020] Determine the coverage range of the base station according to the base station information;

[0021] When the longitude and latitude of the permanent location information are located in the coverage range of the base station, the user behavior information corresponding to the permanent location information is determined as the target coverage user information.

[0022] Optionally, the terminal prediction model includes a price classifier chain model and a brand classifier chain model, and the step of inputting the target coverage user information and the on-sale terminal information into a preset terminal prediction model to generate terminal recommendation information includes:

[0023] Input the target coverage user information into the price classifier chain model to generate a price preference value, and input the target coverage user information into the brand classifier chain model to generate a brand preference value;

[0024] Calculating a price matching value between the price preference value and the terminal information on sale;

[0025] Calculating a brand matching value between the brand preference value and the terminal information on sale;

[0026] Weighted calculation of the price matching value and the brand matching value to generate a terminal matching value;

[0027] The terminal recommendation information is generated according to the size of the terminal matching value.

[0028] Optionally, the price preference value includes a predicted price, and the terminal information on sale includes a terminal on sale price; and the step of calculating a price matching value between the price preference value and the terminal information on sale includes:

[0029] Subtract the predicted price from the selling price of the terminal to generate a price difference;

[0030] Adding one to the absolute value of the price difference to generate a positive difference;

[0031] The reciprocal of the positive difference is determined as the price matching value.

[0032] Optionally, the brand preference value includes a predicted brand identifier, and the terminal information on sale includes a brand identifier on sale; and the step of calculating a brand matching value between the brand preference value and the terminal information on sale includes:

[0033] Determining whether the predicted brand logo matches the brand logo on sale;

[0034] When a match occurs, determining the brand matching value to be a preset characteristic value;

[0035] When there is no match, the brand matching value is determined to be zero.

[0036] In a second aspect of the present invention, an embodiment of the present invention further discloses a terminal recommendation device, including:

[0037] An acquisition module is used to acquire user behavior information, base station information, and terminal information on sale;

[0038] A permanent location information generating module, used to generate permanent location information according to the user behavior information;

[0039] An association module, used to associate the permanent location information with the base station information to determine target coverage user information;

[0040] An input module, used to input the target coverage user information and the on-sale terminal information into a preset terminal prediction model to generate terminal recommendation information;

[0041] The sending module is used to send the terminal recommendation information.

[0042] In the third aspect of the present invention, an embodiment of the present invention further discloses an electronic device, including a processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein the computer program implements the steps of the terminal recommendation method as described above when executed by the processor.

[0043] In a fourth aspect of the present invention, an embodiment of the present invention further discloses a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the terminal recommendation method as described above are implemented.

[0044] The embodiments of the present invention include the following advantages:

[0045] The embodiment of the present invention obtains user behavior information, base station information and terminal information on sale; generates permanent location information based on the user behavior information; associates the permanent location information with the base station information to determine target coverage user information; inputs the target coverage user information and the terminal information on sale into a preset terminal prediction model to generate terminal recommendation information; and sends the terminal recommendation information. The user's permanent location is determined through user behavior data, and users covered by a specific mobile communication network base station are identified in combination with base station information; predictions and recommendations are subsequently made for the covered users, so that the recommended terminals are more in line with user needs; terminal preferences are predicted based on the terminal prediction model to make recommendations, thereby improving the accuracy of user preference recommendations, significantly reducing user selection costs, and improving the accuracy of terminal recommendations. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 is a flowchart of steps of an embodiment of a terminal recommendation method of the present invention;

[0047] Figure 2 is a flowchart of another terminal recommendation method embodiment of the present invention;

[0048] Figure 3 This is a schematic diagram of target coverage users in an embodiment of a terminal recommendation method of the present invention;

[0049] Figure 4 is a preference prediction schematic diagram of an embodiment of a terminal recommendation method of the present invention;

[0050] Figure 5 is a flowchart of an example of a terminal recommendation method of the present invention;

[0051] Figure 6 is a feature fusion schematic diagram of an example of a terminal recommendation method of the present invention;

[0052] Figure 7 It is a structural block diagram of an embodiment of a terminal recommendation device of the present invention. DETAILED DESCRIPTION

[0053] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0054] Reference Figure 1 , shows a flowchart of a terminal recommendation method embodiment of the present invention, which may specifically include the following steps:

[0055] Step 101, obtaining user behavior information, base station information and terminal information on sale;

[0056] In the embodiment of the present invention, the users targeted are all users, that is, all users of mobile terminals used online. Of course, in order to reduce the amount of data processed at a time, all users can also be divided into regions or administrative areas (such as city level), and terminals are recommended based on all users in the region or administrative area.

[0057] Correspondingly, the base station information should also be the information of all base stations in use online. When all users are divided by region or administrative area level, the base station should also be processed using the base station information of the corresponding region or administrative area level. Among them, when recommending a terminal device of a single mobile communication network, the base station information should only obtain the base station information of the base station of the mobile communication network. For example, in an example of the present invention, the recommendation is for a terminal device of a 5G network, and the base station information should be the base station information of all 5G network base stations.

[0058] The terminal information on sale is the information corresponding to all terminals currently on sale. It should be noted that when recommending terminal devices for a single mobile communication network, all terminals refer to all terminals on sale that are applicable to that mobile communication network. For example, when recommending terminal devices for a 5G mobile communication network, the terminal information on sale corresponding to the terminals on sale that are applicable to the 5G network is obtained.

[0059] User-related data can be stored in a user database, base station-related data can be stored in a base station database, and terminal-related data can be stored in a product database. The user database, base station database, and product database can be set on the same physical device (such as a hard board), or on different physical devices; they can also be set on the cloud space; the embodiments of the present invention do not limit this.

[0060] Obtain user behavior information, base station information and terminal information on sale from the corresponding database.

[0061] Step 102, generating permanent location information according to the user behavior information;

[0062] Based on the acquired user behavior information, the behavior of each user is analyzed one by one to find out the geographical location range where each user uses the terminal on a daily basis, and determine the permanent location information of each user. The permanent location information can be classified by date type, such as daily location information for work and daily location information for rest; it can also be classified by time period, such as permanent location information during the day and permanent location information at night.

[0063] Step 103, associating the permanent location information with the base station information to determine target coverage user information;

[0064] The permanent location information of each user is associated with the base station information to determine whether the user's permanent location is within the coverage of the base station, and the users covered by the base station are determined as base station covered users, and the information of base station covered users is determined as target covered user information. In this way, users who can use mobile communication networks such as 5G on a daily basis are screened out. For example, if the base station is a 5G base station, the latitude and longitude data and coverage of the 5G base station are calculated, and the users covered by the 5G base station are identified based on the permanent location information of each user, and the information of the user is determined as the target covered user information.

[0065] Step 104, inputting the target coverage user information and the on-sale terminal information into a preset terminal prediction model to generate terminal recommendation information;

[0066] The preset terminal prediction model can be trained based on sample data and is used to predict the user's preference for the terminal device according to the user's characteristic attributes.

[0067] The target coverage user information and the terminal information on sale can be used as the input data of the model. The terminal prediction model calculates the user's preference data based on the target coverage user information, and compares it with the terminal information on sale to determine the user's preference for the terminals on sale, and generates terminal recommendation information based on the top ranked terminals on sale.

[0068] Step 105: Send the terminal recommendation information.

[0069] After the terminal recommendation information is generated, the terminal recommendation information can be sent to the terminal device corresponding to the user, and the user can view the terminal recommendation information through the terminal device to select a new terminal device. The terminal recommendation information can also be sent to a designated marketing staff, who can recommend terminal devices to the user based on the terminal recommendation information.

[0070] The embodiment of the present invention obtains user behavior information, base station information and terminal information on sale; generates permanent location information based on the user behavior information; associates the permanent location information with the base station information to determine target coverage user information; inputs the target coverage user information and the terminal information on sale into a preset terminal prediction model to generate terminal recommendation information; and sends the terminal recommendation information. The user's permanent location is determined through user behavior data, and users covered by a specific mobile communication network base station are identified in combination with base station information; predictions and recommendations are subsequently made for the covered users, so that the recommended terminals are more in line with user needs; terminal preferences are predicted based on the terminal prediction model to make recommendations, thereby improving the accuracy of user preference recommendations, significantly reducing user selection costs, and improving the accuracy of terminal recommendations.

[0071] Reference Figure 2 , shows a flowchart of another terminal recommendation method embodiment of the present invention, which may specifically include the following steps:

[0072] Step 201, obtaining user behavior information, base station information and terminal information on sale;

[0073] From the specified database, obtain the user's user package information, location information, number of times the App (Application) is accessed from another network, current terminal information, traffic usage information, etc. as user behavior information.

[0074] The base station information of the mobile communication network is obtained from the specified database. Such as the base station of the 5G mobile communication network, the base station of the 6G mobile communication network, etc. The embodiment of the present invention takes the 5G mobile communication network as an example to illustrate the embodiment of the present invention, but is not limited to the 5G mobile communication network. Specifically, the longitude and latitude data and coverage data of the 5G base station are obtained as the base station information.

[0075] The configuration information, price, brand and other information of all the terminals on sale applicable to the 5G network are obtained as the terminal on sale information. The terminals on sale may be all the terminals on sale in the market, or the terminals on sale of users in a third-party enterprise such as a dealer.

[0076] Step 202, generating permanent location information according to the user behavior information;

[0077] According to the user behavior information, the top several geographical locations where the user resides are screened out to generate the permanent location information. For example, according to the user behavior information, the top three geographical locations where the user resides are screened out to generate the permanent location information.

[0078] In one embodiment of the present invention, the user behavior information includes location information and traffic usage information, and the step of generating permanent location information based on the user behavior information includes:

[0079] It should be noted that the location information is the location information corresponding to the user's stay location in the historical time period, and the traffic usage information is the traffic usage of the user in the historical time period. The historical time period can be determined according to demand, such as the past month, the past three months, the past six months, etc.; the embodiment of the present invention does not specifically limit this.

[0080] Sub-step S2021, for any location information, determining the longitude and latitude data corresponding to the location information;

[0081] For a single user, for any location information, the latitude and longitude data for the location are determined from the historical location information of the user to determine the geographical location corresponding to the location information.

[0082] Sub-step S2022, classifying the traffic usage information to generate the traffic usage on weekdays and the traffic usage on weekends;

[0083] For a single user, the user's traffic usage information is classified according to the date of traffic usage, and divided into weekday traffic usage and weekend traffic usage.

[0084] Sub-step S2023, associating the weekday traffic usage with the latitude and longitude data to generate weekday location data, and associating the rest day traffic usage with the latitude and longitude data to generate rest day location data;

[0085] The weekday traffic usage is associated with the corresponding longitude and latitude data to determine the location where the user uses traffic on weekdays, and generate the weekday location data corresponding to the location where the user uses traffic on weekdays. The weekday location data is used to characterize the geographical location where the user uses traffic on weekdays.

[0086] The rest day traffic usage is associated with the corresponding longitude and latitude data to determine the location where the user uses traffic on the rest day, and generate rest day location data corresponding to the location where the user uses traffic on the rest day. Rest day location data is used to characterize the geographical location where the user uses traffic on the rest day. It should be noted that the total number of working days and rest days is consistent with the total number of days in the historical time period. If the user location information and traffic usage information of the past 30 days are used, there are 22 working days and 8 rest days.

[0087] Sub-step S2024, sorting the working day location data according to the size of the working day traffic usage to generate working day residence data; sorting the rest day location data according to the size of the rest day traffic usage to generate rest day residence data;

[0088] Since the workday location data and the rest day location data correspond to multiple geographic locations respectively, the top several geographic locations with the highest traffic usage can be extracted according to the traffic usage of the users in the geographic locations, and the workday location data and rest day location data can be generated. For example, the top three users with the highest traffic usage corresponding to the workday location data are the workday location data, and the top three users with the highest traffic usage corresponding to the rest day location data are the rest day location data.

[0089] Specifically, take the first three as an example to illustrate:

[0090] The generated multiple working day location data can be sorted from large to small according to the traffic usage, and the working day location data corresponding to the top three traffic usages can be merged as the working day location data. The generated multiple rest day location data can be sorted from large to small according to the traffic usage, and the rest day location data corresponding to the top three traffic usages can be merged as the rest day location data.

[0091] Sub-step S2025, determining the work daily location data and the rest daily location data as permanent location information.

[0092] After obtaining the daily work location data and the daily rest location data, all geographical locations corresponding to the daily work location data and the daily rest location data can be used as permanent location information. Among them, the geographical locations corresponding to the daily work location data and the daily rest location data can be represented by longitude and latitude data.

[0093] It should be noted that when all geographical locations corresponding to the work daily residence data and all geographical locations corresponding to the rest daily residence data overlap, the duplicate parts should be deleted to avoid redundancy of geographical locations in the permanent residence information.

[0094] For example, when the latitude and longitude data of the geographical location corresponding to the work daily station data are A, B, and C, and the latitude and longitude data of the geographical location corresponding to the rest daily station data are D, E, and F, the geographical location of the permanent station information includes A, B, C, D, E, and F. When the latitude and longitude data of the geographical location corresponding to the work daily station data are A, B, and C, and the latitude and longitude data of the geographical location corresponding to the rest daily station data are A, D, and E, the geographical location of the permanent station information includes A, B, C, D, and E.

[0095] Step 203, associating the permanent location information with the base station information to determine target coverage user information;

[0096] The permanent location information and 5G base station information are associated to filter out users that can be covered within the existing 5G base station coverage, and determine the target coverage user information of the user.

[0097] In one embodiment of the present invention, the step of associating the resident location information with the base station information to determine target coverage user information includes:

[0098] Sub-step S2031, determining the base station coverage range according to the base station information;

[0099] The base station information may record the longitude and latitude of the base station, the signal coverage radius of the base station, and a circular area formed with the longitude and latitude of the base station as the center and the signal coverage radius as the radius of the circle. The longitude and latitude range covered by the circular area is the base station coverage range.

[0100] It should be noted that, when there are two or more base stations and the circular areas corresponding to the two base stations overlap, the longitude and latitude range covered by the fan-shaped areas corresponding to the base stations may be used as the base station coverage range.

[0101] Sub-step S2032: when the longitude and latitude of the permanent location information are within the coverage range of the base station, determining the user behavior information corresponding to the permanent location information as target coverage user information.

[0102] The longitude and latitude of the permanent location information may be compared with the base station coverage, and the comparison method includes but is not limited to the mapping method and the difference method. When the longitude and latitude of the permanent location information are within the longitude and latitude range corresponding to the base station coverage range, the user behavior information corresponding to the permanent location information is determined as the target coverage user information. In addition, when the longitude and latitude of the permanent location information are outside the longitude and latitude range corresponding to the base station coverage range, the user behavior information corresponding to the permanent location information may be filtered out.

[0103] You can refer to Figure 3 , showing a schematic diagram of target coverage users in an embodiment of a terminal recommendation method of the present invention,

[0104] Taking the longitude and latitude of the 5G base station as the center and the signal coverage radius of the base station as the circle, the sector area surrounded by the two base stations is the coverage range of the 5G base station, and users A, B, C, and E are in the coverage range of the 5G base station as 5G base station coverage users. The user behavior data of users A, B, C, and E can be determined as target coverage user information. Users D and F are not 5G base station coverage users, and their user behavior information can be filtered out.

[0105] Step 204, inputting the target coverage user information and the on-sale terminal information into a preset terminal prediction model to generate terminal recommendation information;

[0106] The target coverage user information and the terminal information on sale are input into the preset terminal prediction model, the matching degree between the target coverage user and the terminal on sale is calculated, and the terminal recommendation information is generated based on the terminal on sale with a higher matching degree.

[0107] In one embodiment of the present invention, the terminal prediction model includes a price classifier chain model and a brand classifier chain model; it should be noted that before predicting the terminals on sale, the price classifier chain model and the brand classifier chain model can be established and trained based on sample data. The user behavior data can be split into price attribute data and brand attribute data, and a binary association algorithm (multi-label classification model) is used to train the price attribute data as sample data to obtain a price classifier chain model; a binary association algorithm is used to train the brand attribute data as sample data to obtain a brand classifier chain model. The price classifier chain model is used to predict the user's price preference, and the brand classifier chain model is used to predict the user's brand preference. And the output results of the two models can be fused to determine the user's preference attributes for the terminal.

[0108] The step of inputting the target coverage user information and the on-sale terminal information into a preset terminal prediction model to generate terminal recommendation information includes:

[0109] Sub-step S2041, inputting the target coverage user information into the price classifier chain model to generate a price preference value, and inputting the target coverage user information into the brand classifier chain model to generate a brand preference value;

[0110] The target coverage user information can be input into the price classifier chain model, and the price classification model is fitted according to the price attribute characteristics of the target coverage user to output a price preference value, which can specifically be a predicted price.

[0111] The target coverage user information can be input into the brand classifier chain model, and the price classification model is fitted according to the brand attribute characteristics of the target coverage user to output a brand preference value. The brand preference value can specifically be a brand identity such as a brand name.

[0112] Sub-step S2042, calculating a price matching value between the price preference value and the terminal information on sale;

[0113] The obtained price preference value is compared with the current selling price of the terminal in the terminal information on sale, and the price matching value between each terminal on sale and the user is calculated.

[0114] In one embodiment of the present invention, the price preference value includes a predicted price, and the terminal information on sale includes a terminal on sale price; and the step of calculating a price matching value between the price preference value and the terminal information on sale includes:

[0115] Sub-step S20421, subtracting the predicted price from the selling price of the terminal to generate a price difference;

[0116] In practical applications, the terminal selling price minus the predicted price can be used to generate the price difference between the current terminal selling price and the predicted price.

[0117] Sub-step S20422, adding one to the absolute value of the price difference to generate a positive difference;

[0118] The absolute value of the price difference can be taken so that the price difference can be compared in the same direction and the price comparison direction can be unified. In addition, the absolute value of the price difference can be added by one to generate a non-zero positive difference.

[0119] Sub-step S20423, determining the reciprocal of the positive difference as the price matching value.

[0120] In practical applications, the reciprocal of the positive difference can be used as the price matching value. Specifically, the positive difference data can be used as the denominator and the numerator can be used as one to calculate the price matching value.

[0121] Sub-step S2043, calculating the brand matching value between the brand preference value and the terminal information on sale;

[0122] Similarly, the obtained brand preference value is compared with the current selling price of the terminal brand in the terminal information on sale, and the price matching value between each terminal on sale and the user is calculated.

[0123] In one embodiment of the present invention, the brand preference value includes a predicted brand identifier, and the terminal information on sale includes a brand identifier on sale; and the step of calculating a brand matching value between the brand preference value and the terminal information on sale includes:

[0124] Sub-step S20431, determining whether the predicted brand logo matches the brand logo on sale;

[0125] In actual application, the predicted brand logo and the brand logo on sale can be matched one by one to determine whether the predicted brand logo matches the brand logo on sale. Those skilled in the art can select different consistency matching methods according to requirements, and the embodiments of the present invention do not specifically limit this.

[0126] Sub-step S20432, when matching, determining the brand matching value as a preset feature value;

[0127] When the predicted brand identifier is consistent with any of the on-sale brand identifiers, it is determined that the predicted brand identifier matches the on-sale brand identifier, and the preset feature value is determined as the brand matching value. The preset feature value can be set according to different calculation methods. For example, in an example of the present invention, the feature value can be one; that is, when the predicted brand identifier matches the on-sale brand identifier, the brand matching value is one.

[0128] Sub-step S20433, when there is no match, determining that the brand matching value is zero.

[0129] When the predicted brand logo is inconsistent with the logo in the brand logo on sale, it is determined that the predicted brand logo does not match the brand logo on sale, and the brand matching value may be determined to be zero.

[0130] Sub-step S2044, weighted calculation of the price matching value and the brand matching value to generate a terminal matching value;

[0131] After obtaining the price matching value and the brand matching value, the corresponding weight values ​​can be assigned to the price matching value and the brand matching value, and the price matching value and the brand matching value can be weighted calculated after the weight values ​​are assigned, and the calculation result is the terminal matching value. That is, the terminal matching value can be calculated using the following formula:

[0132]

[0133]

[0134] Among them, α1 and α2 are price weight coefficient and brand weight coefficient respectively, α1+α2=1;

[0135] Y1 is the selling price, and Y2 is the predicted price;

[0136] P is the brand matching value.

[0137] For example, see Figure 4 , showing a preference prediction schematic diagram of a terminal recommendation method embodiment of the present invention, the price features such as the monthly outgoing amount and the recharge amount in the user behavior information can be input into the price classifier chain model to generate a price preference value. The brand features such as the historical terminal brand and the commonly used APP preference in the user behavior information can be input into the brand classifier chain model to generate a brand preference value. The price preference value and the terminal information on sale are weighted to calculate the terminal matching value.

[0138] Sub-step S2045: generating terminal recommendation information according to the size of the terminal matching value.

[0139] In an embodiment of the present invention, after calculating the terminal matching value of each terminal on sale, the terminals on sale can be sorted in descending order according to the size of the terminal matching value of each terminal on sale, and the terminals on sale corresponding to the top several ranked terminals are taken, and the terminal recommendation information is generated in combination with the preferential package information, coupon information, purchase method and other information corresponding to the terminals on sale.

[0140] Step 205, sending the terminal recommendation information;

[0141] The terminal recommendation information is sent to the mobile terminal currently used by the user, and the mobile terminal currently used by the user displays the terminal recommendation information for the user to view.

[0142] Step 206, obtaining user transaction information;

[0143] When a user purchases a new 5G mobile terminal, the user transaction information corresponding to the user's 5G mobile terminal can be obtained, where the user transaction information includes but is not limited to the price, brand, model, time, and coupon redemption status of the purchase of the 5G terminal.

[0144] Step 207: adjusting the model parameters of the terminal prediction model based on the user transaction information.

[0145] The user transaction information is used as positive sample data and the model parameters of the terminal prediction model to modify the terminal prediction model and improve the prediction accuracy of the terminal prediction model. In addition, the preferential plan can be adjusted according to the information of the user using the coupon in the user transaction information, so that the terminal recommendation information is more complete and accurate.

[0146] The embodiments of the present invention determine the user's permanent location through user behavior data, and identify users covered by a specific mobile communication network base station in combination with base station information; subsequently, predictive recommendations are made for the covered users, so that the recommended terminals are more in line with user needs; terminal prediction models are used to predict user preferences for terminals and make recommendations, thereby improving the accuracy of user preference recommendations, significantly reducing user selection costs, and improving the accuracy of terminal recommendations; and different classifier chain models are used for predictive identification based on the price attributes and brand attributes of 5G terminals, and while identifying potential 5G terminal users, the preferences for each attribute are predicted separately, and then the prediction results of each model are combined to obtain the 5G terminal to be recommended, thereby significantly improving the prediction accuracy.

[0147] Reference Figure 5 , shows a flowchart of an example of a terminal recommendation method of the present invention, which may specifically include the following steps:

[0148] Step S1, obtaining user information (such as location information, terminal information, etc.), 5G base station information and information of 5G terminals on sale;

[0149] Obtain user information, 5G base station information, and 5G terminal information on sale. Among them, user information includes: user package information, location information, number of visits to apps on other networks, current terminal information, etc. 5G base station information includes 5G base station location and coverage radius, and 5G terminal information on sale includes 5G terminal brand and price. Specifically, the above data can be obtained by cleaning and feature processing the data of each operator's system. For example, refer to Figure 6, the acquired mobile DPI (Deep packet inspection) data, CRM (Customer Relationship Management) acceptance data, billing system data, 5G base station data, and terminal platform data can be feature fused to obtain personal physical sign indicators, business characteristic indicators, terminal preference indicators, base station location indicators, and 5G terminal indicators; based on these characteristics, price financial data such as user age, family composition, terminal price, package price, consumption trend, average monthly bill amount, average monthly recharge amount, etc., such as user location information such as the user's permanent longitude and latitude during the day on weekdays, permanent longitude and latitude at night on weekdays, permanent longitude and latitude during the day on weekends, and permanent longitude and latitude at night on weekends, as well as 5G base station location information, configuration information, sector location, coverage, 5G terminal brand, price and other base station information and terminal information on sale.

[0150] Step S2, associating the user location information with the 5G base station information to identify users covered by the 5G base station;

[0151] Based on the user's location information and traffic usage, the three-day longitude and latitude data on weekdays and the three-day longitude and latitude data on weekends are identified as permanent location data. Combined with the longitude and latitude data and coverage range of the 5G base station, the target coverage users covered by the 5G base station are identified.

[0152] Step S3, establish a 5G terminal prediction model to identify the 5G terminal preferred by the above user

[0153] The matching scores of the predicted 5G terminal price and brand and the prices and brands of 5G terminals currently on sale are calculated and the top three 5G terminals on sale with the highest matching scores are selected to generate terminal recommendation information.

[0154] Step S4: Send the recommended 5G terminal preferential information to the user's mobile phone

[0155] The price information, coupon information, and contact information of the recommended 5G terminal are sent to the user's mobile phone.

[0156] Step S5: Obtain the user's 5G terminal transaction information

[0157] Obtain users' 5G terminal transaction information and coupon redemption, including the price, brand, model, and time of purchasing the 5G terminal.

[0158] Step S6: Effect feedback and system optimization

[0159] Based on the user's 5G terminal transaction data, adjust the model parameters and marketing information in S1 to S4 above to optimize the system effect.

[0160] It should be noted that, for the sake of simplicity, the method embodiments are described as a series of action combinations, but those skilled in the art should be aware that the embodiments of the present invention are not limited by the order of the actions described, because according to the embodiments of the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of the present invention.

[0161] Reference Figure 7 , shows a structural block diagram of an embodiment of a terminal recommendation device of the present invention, which may specifically include the following modules:

[0162] The acquisition module 701 is used to acquire user behavior information, base station information and terminal information on sale;

[0163] A permanent location information generating module 702, configured to generate permanent location information according to the user behavior information;

[0164] An association module 703 is used to associate the permanent location information with the base station information to determine target coverage user information;

[0165] An input module 704 is used to input the target coverage user information and the on-sale terminal information into a preset terminal prediction model to generate terminal recommendation information;

[0166] The sending module 705 is used to send the terminal recommendation information.

[0167] In one embodiment of the present invention, the device further comprises:

[0168] Transaction information module, used to obtain user transaction information;

[0169] A feedback module is used to adjust the model parameters of the terminal prediction model based on the user transaction information.

[0170] In one embodiment of the present invention, the user behavior information includes location information and traffic usage information, and the permanent location information generation module 702 includes:

[0171] A user longitude and latitude determination submodule, used to determine the longitude and latitude data corresponding to any location information;

[0172] A classification submodule, used to classify the traffic usage information and generate the traffic usage on weekdays and the traffic usage on weekends;

[0173] An associating submodule, used to associate the working day traffic usage with the latitude and longitude data to generate working day location data, and associate the rest day traffic usage with the latitude and longitude data to generate rest day location data;

[0174] A sorting submodule is used to sort the working day location data according to the size of the working day traffic usage to generate working day resident data; sort the rest day location data according to the size of the rest day traffic usage to generate rest day resident data;

[0175] The permanent location information determination submodule is used to determine the daily work location data and the daily rest location data as permanent location information.

[0176] In one embodiment of the present invention, the association module 703 includes:

[0177] A base station coverage range determination submodule, used to determine the base station coverage range according to the base station information;

[0178] The coverage determination submodule is used to determine the user behavior information corresponding to the permanent location information as target coverage user information when the longitude and latitude of the permanent location information are within the coverage range of the base station.

[0179] In one embodiment of the present invention, the terminal prediction model includes a price classifier chain model and a brand classifier chain model, and the input module 704 includes:

[0180] An input submodule, used for inputting the target coverage user information into the price classifier chain model to generate a price preference value, and inputting the target coverage user information into the brand classifier chain model to generate a brand preference value;

[0181] A first calculation submodule, used for calculating a price matching value between the price preference value and the terminal information on sale;

[0182] A second calculation submodule is used to calculate a brand matching value between the brand preference value and the terminal information on sale;

[0183] A weighted calculation submodule, used for weighted calculation of the price matching value and the brand matching value to generate a terminal matching value;

[0184] The terminal recommendation information determination submodule is used to generate terminal recommendation information according to the size of the terminal matching value.

[0185] In one embodiment of the present invention, the price preference value includes a predicted price, the terminal information on sale includes a terminal on sale price; and the first calculation submodule includes:

[0186] a difference making unit, used for subtracting the predicted price from the selling price of the terminal to generate a price difference;

[0187] an absolute value unit, used to add one to the absolute value of the price difference to generate a positive difference;

[0188] The reciprocal unit is used to determine the reciprocal of the positive difference as the price matching value.

[0189] In one embodiment of the present invention, the brand preference value includes a predicted brand identifier, and the on-sale terminal information includes an on-sale brand identifier; the second calculation submodule includes:

[0190] A judging unit, configured to judge whether the predicted brand logo matches the brand logo on sale;

[0191] a first brand matching value determining unit, configured to determine the brand matching value as a preset characteristic value when the predicted brand identifier matches the on-sale brand identifier;

[0192] The second brand matching value determining unit is configured to determine that the brand matching value is zero when the predicted brand identifier does not match the brand identifier on sale.

[0193] As for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0194] An embodiment of the present invention further provides an electronic device, including:

[0195] A processor and a storage medium, wherein the storage medium stores a computer program executable by the processor, and when the electronic device is running, the processor executes the computer program to perform the method as described in any one of the embodiments of the present invention. The specific implementation method and technical effect are similar to those of the method embodiment, and will not be repeated here.

[0196] The embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, and the computer program is executed by a processor to execute the method as described in any one of the embodiments of the present invention. The specific implementation method and technical effect are similar to those of the method embodiment, and will not be repeated here.

[0197] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0198] It will be appreciated by those skilled in the art that the embodiments of the present invention may be provided as methods, devices, or computer program products. Therefore, the embodiments of the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the embodiments of the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0199] The embodiments of the present invention are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0200] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0201] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0202] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.

[0203] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0204] The terminal recommendation method, the terminal recommendation device, the electronic device and the storage medium provided by the present invention are introduced in detail above. The principle and implementation mode of the present invention are explained by using specific examples in this article. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation mode and the scope of application. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A terminal recommendation method, characterized in that: include: Obtain user behavior information, base station information, and terminal information on sale; Generate permanent location information based on the user behavior information; Associating the permanent location information with the base station information to determine target coverage user information; Inputting the target coverage user information and the on-sale terminal information into a preset terminal prediction model to generate terminal recommendation information; Sending the terminal recommendation information; The step of associating the resident location information with the base station information to determine target coverage user information includes: Determine the coverage range of the base station according to the base station information; When the longitude and latitude of the permanent location information are within the coverage range of the base station, determining the user behavior information corresponding to the permanent location information as target coverage user information; The terminal prediction model includes a price classifier chain model and a brand classifier chain model, and the step of inputting the target coverage user information and the on-sale terminal information into a preset terminal prediction model to generate terminal recommendation information includes: Input the target coverage user information into the price classifier chain model to generate a price preference value, and input the target coverage user information into the brand classifier chain model to generate a brand preference value; Calculating a price matching value between the price preference value and the terminal information on sale, wherein the price preference value includes a predicted price, the terminal information on sale includes a terminal on sale price, and the price matching value is obtained by calculating the difference between the predicted price and the terminal on sale price; Calculating a brand matching value between the brand preference value and the terminal information on sale, wherein the brand preference value includes a predicted brand identifier, the terminal information on sale includes a brand identifier on sale, and the brand matching value is obtained by matching the predicted brand identifier with the brand identifier on sale; Weighted calculation of the price matching value and the brand matching value to generate a terminal matching value; The terminal recommendation information is generated according to the size of the terminal matching value.

2. The method according to claim 1, characterized in that The method further comprises: Obtain user transaction information; The model parameters of the terminal prediction model are adjusted based on the user transaction information.

3. The method according to claim 1 or 2, characterized in that: The user behavior information includes location information and traffic usage information. The step of generating permanent location information based on the user behavior information includes: For any location information, determine the longitude and latitude data corresponding to the location information; Classify the traffic usage information to generate traffic usage on weekdays and traffic usage on weekends; Associating the working day traffic usage with the longitude and latitude data to generate working day location data, and associating the rest day traffic usage with the longitude and latitude data to generate rest day location data; The working day location data are sorted according to the size of the working day traffic usage to generate working day residence data; the rest day location data are sorted according to the size of the rest day traffic usage to generate rest day residence data; The working daily location data and the resting daily location data are determined as permanent location information.

4. The method according to claim 1, characterized in that The step of calculating the price matching value between the price preference value and the terminal information on sale includes: Subtract the predicted price from the selling price of the terminal to generate a price difference; Adding one to the absolute value of the price difference to generate a positive difference; The reciprocal of the positive difference is determined as the price matching value.

5. The method according to claim 1, characterized in that The step of calculating the brand preference value and the brand matching value of the terminal information on sale includes: Determining whether the predicted brand logo matches the brand logo on sale; When a match occurs, determining the brand matching value to be a preset characteristic value; When there is no match, the brand matching value is determined to be zero.

6. A terminal recommendation device, characterized in that: include: An acquisition module is used to acquire user behavior information, base station information, and terminal information on sale; A permanent location information generating module, used to generate permanent location information according to the user behavior information; An association module, used to associate the permanent location information with the base station information to determine target coverage user information; An input module, used to input the target coverage user information and the on-sale terminal information into a preset terminal prediction model to generate terminal recommendation information; A sending module, used for sending the terminal recommendation information; Wherein, the association module includes: A base station coverage range determination submodule, used to determine the base station coverage range according to the base station information; A coverage determination submodule, configured to determine, when the longitude and latitude of the permanent location information are within the coverage range of the base station, the user behavior information corresponding to the permanent location information as target coverage user information; The terminal prediction model includes a price classifier chain model and a brand classifier chain model, and the input module includes: An input submodule, used for inputting the target coverage user information into the price classifier chain model to generate a price preference value, and inputting the target coverage user information into the brand classifier chain model to generate a brand preference value; A first calculation submodule, configured to calculate a price matching value between the price preference value and the terminal information on sale, wherein the price preference value includes a predicted price, the terminal information on sale includes a terminal on sale price, and the price matching value is obtained by calculating the difference between the predicted price and the terminal on sale price; A second calculation submodule is used to calculate a brand matching value between the brand preference value and the terminal information on sale, wherein the brand preference value includes a predicted brand identifier, the terminal information on sale includes a brand identifier on sale, and the brand matching value is obtained by matching the predicted brand identifier with the brand identifier on sale; A weighted calculation submodule, used for weighted calculation of the price matching value and the brand matching value to generate a terminal matching value; The terminal recommendation information determination submodule is used to generate terminal recommendation information according to the size of the terminal matching value.

7. An electronic device, characterized in that: The invention comprises a processor, a memory and a computer program stored in the memory and capable of running on the processor, wherein the computer program implements the steps of the terminal recommendation method according to any one of claims 1 to 5 when executed by the processor.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the terminal recommendation method according to any one of claims 1 to 5 are implemented.

Images