Broadband marketing strategy making method and device and readable storage medium

Abstract

The invention provides a broadband marketing strategy making method and device and a readable storage medium, and relates to the technical field of communication, and the method comprises the steps: obtaining EDPI data corresponding to target broadband account information in a preset time interval; obtaining a poor network quality score of the user in each time period in a preset time interval according to the EDPI data; according to the poor network quality score of each time period, obtaining a network quality change condition of the user in a preset time interval, and obtaining a poor network quality portrait of the user; and performing user family attribute portrait analysis based on the EDPI data to obtain a key dimension for constructing a user family attribute accurate portrait. According to the method, the device and the medium, the problem that an existing broadband poor quality detection technology and a user portrait construction method can only help an operator to identify and solve network problems to a certain extent and cannot be effectively converted into a marketing strategy for improving user satisfaction and enhancing competitiveness can be solved.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a method, device and readable storage medium for formulating a broadband marketing strategy. Background Art

[0002] As competition in the communications industry intensifies, operators' traditional user expansion strategies are gradually giving way to in-depth operations of existing users, improving service quality, and exploring new profit models through value-added services. This shift requires operators to make fundamental adjustments in their strategies to cope with the increasingly complex and diversified customer needs. Traditional marketing methods, such as physical promotion and telephone marketing, are not only inefficient and costly, but also often lead to a mismatch between marketing content and actual consumer needs due to the lack of detailed analysis of the target market. In this context, the rise of artificial intelligence and big data technologies has provided operators with new opportunities to help them better understand and meet customer needs, making precision marketing possible.

[0003] At present, although AI (Artificial Intelligence) and big data technologies provide strong support for precision marketing in theory, they are still limited in practical applications, especially in the accurate identification and service recommendation of home user groups. In addition, the rapid popularization and speed increase of broadband networks have also brought new challenges. Users' requirements for network quality are constantly increasing, and poor broadband quality has become a key factor affecting user experience.

[0004] However, although the existing broadband quality detection technology and user profile construction methods can help operators identify and solve network problems to a certain extent, they are not well integrated and cannot be effectively transformed into marketing strategies to improve user satisfaction and enhance competitiveness. Therefore, innovative marketing technologies and strategies are urgently needed to achieve more accurate market positioning, higher operational efficiency and lower marketing costs. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a method, device and readable storage medium for formulating a broadband marketing strategy in response to the above-mentioned shortcomings of the prior art, so as to solve the problem that the existing broadband quality difference detection technology and user portrait construction method can only help operators identify and solve network problems to a certain extent, but cannot be effectively transformed into a marketing strategy to improve user satisfaction and enhance competitiveness.

[0006] In a first aspect, the present invention provides a method for formulating a broadband marketing strategy,

[0007] Methods include:

[0008] Obtaining flexible packet inspection technology EDPI data corresponding to target broadband account information within a preset time interval;

[0009] Obtaining a network quality score of the user in each time period within the preset time interval according to the EDPI data;

[0010] According to the network quality score of each time period, the network quality change of the user in the preset time interval is obtained to obtain the network quality profile of the user;

[0011] Performing a user family attribute profile analysis based on the EDPI data to obtain key dimensions for constructing an accurate user family attribute profile, wherein the key dimensions include at least two of the user's Internet preferences, family consumption capacity and resident population, family member age distribution, and active time period;

[0012] Constructing an accurate portrait of the user's family attributes according to the key dimensions used to construct an accurate portrait of the user's family attributes; and,

[0013] A broadband marketing strategy is formulated based on the network quality profile and the accurate profile of the user's family attributes.

[0014] Furthermore, obtaining the network quality score of the user in each time period within the preset time interval according to the EDPI data specifically includes:

[0015] According to the EDPI data of each time period corresponding to the target broadband account information, the network performance index of each time period is obtained, wherein the network performance index includes the number of ACK packets, the number of packet losses, the first packet delay and the overall delay;

[0016] The network quality score of the user in each time period is calculated according to the number of ACK packets, the number of packet losses, the first packet delay and the overall delay in each time period.

[0017] Further, the calculating of the network quality score of the user in each time period according to the number of ACK packets, the number of packet losses, the first packet delay and the overall delay in each time period specifically includes:

[0018] Calculate the packet loss rate of each time period according to the number of ACK packets and the number of packet losses in each time period;

[0019] The network quality score for each time period is calculated according to the following formula:

[0020]

[0021] In the formula, Score pQ Indicates the network quality score, Score PLR Indicates the packet loss rate, Score FBL Indicates the first packet delay, Score OLrepresents the overall delay, and α represents the correction factor.

[0022] Furthermore, the user family attribute portrait analysis is performed based on the EDPI data to obtain key dimensions for constructing an accurate user family attribute portrait, including at least two of the following:

[0023] Obtaining a standardized application program APP name and traffic type based on the application, access domain name and uniform resource locator URL path information in the EDPI data, and statistically analyzing the usage time ratio of each type of APP through the standardized APP name and traffic type to obtain the user's Internet preferences;

[0024] Determine the specific terminal devices and quantity used in the user's home according to the Hypertext Transfer Protocol HTTP header request information and the domain name in the EDPI data, and obtain the user's home consumption capacity and resident population based on the specific terminal devices and quantity used in the user's home;

[0025] Identify whether there are children and elderly people in the user's family based on the EDPI data, and obtain the age distribution of the user's family members;

[0026] The active time period of each family member in the user's home is analyzed according to the EDPI data, and the preset autoregressive moving average model ARMA model is used to predict the active time period of each family member at home in the future time interval.

[0027] Further, the determining of the specific terminal devices and the number of the terminal devices used in the user's home according to the Hypertext Transfer Protocol HTTP header request information and the domain name in the EDPI data specifically includes:

[0028] Extracting a terminal identifier from the HTTP header request information and the domain name in the EDPI data;

[0029] Compare the extracted terminal identifier with the pre-established terminal information database to obtain the specific terminal devices and quantity used in the user's home;

[0030] The obtaining of the household consumption capacity and resident population of the user based on the specific terminal devices and the number of terminal devices used in the household of the user specifically includes:

[0031] Calculate a comprehensive score of the potential consumption concept of the user's family based on the specific terminal devices and quantity used in the user's family;

[0032] The potential consumption concept of the user's family is obtained by comprehensively scoring the potential consumption concept of the user's family;

[0033] Calculate the potential consumption level score of the user's family based on the potential consumption concept of the user's family and the broadband package information of the user's family;

[0034] The potential consumption level of the user's family is obtained according to the potential consumption level score of the user's family;

[0035] Identify the size of the permanent population of the user's home based on the active time period of each terminal device among the specific terminal devices used in the user's home;

[0036] According to the broadband installation address of the user's family, the residential property of the user's family is obtained with the help of a real estate information platform, and the residential consumption level of the user's family is obtained based on the residential property.

[0037] Furthermore, the comprehensive score of the potential consumption concept of the user's family is calculated based on the specific terminal devices and the number of terminal devices used in the user's family, and is specifically calculated according to the following formula:

[0038]

[0039] In the formula, C represents the comprehensive score of the user's family's potential consumption concept, n represents the number of specific terminal devices used in the user's family, and w i represents the weight of the i-th terminal device, C i represents the household consumption concept reflected by the i-th terminal device;

[0040] The potential consumption level score of the user's family is calculated based on the potential consumption concept of the user's family and the broadband package information of the user's family, and is specifically calculated according to the following formula:

[0041] L=w C ×C+w p ×P

[0042] Where L represents the potential consumption level score of the user's family, w C represents the weight of the comprehensive score of potential consumption concept, P represents the score of the broadband package information of the user's family, and w p Indicates the weight of the broadband package information score.

[0043] Furthermore, the broadband package information score P of the user's family is calculated according to the following formula:

[0044] P=α×S p +β×S s +γ×S d

[0045] In the formula, s p Indicates the monthly fee of the broadband package, S s Indicates the number of additional services in the broadband package, S d represents the package period of the broadband package, α, β, and γ represent Sp , S s , S d The weight it occupies.

[0046] Furthermore, the identifying the size of the resident population of the user's home based on the active time period of each terminal device among the specific terminal devices used in the user's home specifically includes:

[0047] Taking terminal devices as units, count the active time periods of each terminal device in the user's home, and calculate the average usage time ratio of each terminal device within the preset time intervals of various APP applications;

[0048] According to the active time period of each terminal device and the average usage time ratio of each terminal device within the preset time interval of each type of APP application, the usage patterns of the terminal devices are clustered to obtain the permanent population size of the user's family.

[0049] In a second aspect, the present invention provides a device for formulating a broadband marketing strategy, the device comprising:

[0050] A data acquisition module, used to acquire flexible packet inspection technology EDPI data corresponding to target broadband account information within a preset time interval;

[0051] A quality difference score acquisition module, connected to the data acquisition module, for acquiring the network quality difference score of the user in each time period within the preset time interval according to the EDPI data;

[0052] A quality difference portrait acquisition module, connected to the quality difference score acquisition module, is used to obtain the network quality change of the user in the preset time interval according to the network quality difference score of each time period, and obtain the network quality difference portrait of the user;

[0053] A key dimension acquisition module, connected to the data acquisition module, for performing a user family attribute portrait analysis based on the EDPI data to obtain key dimensions for constructing an accurate user family attribute portrait, wherein the key dimensions include at least two of the user's Internet preferences, family consumption capacity and resident population, family member age distribution, and active time period;

[0054] A family portrait construction module, connected to the key dimension acquisition module, for constructing an accurate portrait of the user's family attributes according to the key dimensions used to construct an accurate portrait of the user's family attributes;

[0055] The marketing strategy formulation module is connected to the quality difference portrait acquisition module and the family portrait construction module, and is used to formulate a broadband marketing strategy based on the network quality difference portrait and the accurate portrait of the user's family attributes.

[0056] In a third aspect, the present invention provides a broadband marketing strategy formulation device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to implement the broadband marketing strategy formulation method described in the first aspect.

[0057] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the method for formulating a broadband marketing strategy described in the first aspect above is implemented.

[0058] The present invention provides a method, device and readable storage medium for formulating broadband marketing strategies. First, obtain the flexible packet detection technology EDPI data corresponding to the target broadband account information within a preset time interval; then obtain the network quality difference score of the user in each time period within the preset time interval based on the EDPI data; and obtain the network quality change of the user in the preset time interval based on the network quality difference score of each time period, and obtain the network quality difference portrait of the user; at the same time, perform user family attribute portrait analysis based on the EDPI data to obtain key dimensions for constructing an accurate portrait of the user's family attributes, the key dimensions include at least two of the user's Internet preferences, family consumption capacity and resident population, family member age distribution, and active time periods; and construct an accurate portrait of the user's family attributes based on the key dimensions for constructing an accurate portrait of the user's family attributes; finally, formulate a broadband marketing strategy based on the network quality difference portrait and the accurate portrait of the user's family attributes. The present invention can simultaneously obtain a network quality difference portrait and an accurate portrait of the user's family attributes through the analysis of EDPI data. Since the combination of the two can provide more comprehensive user characteristics, more accurate and personalized marketing strategies can be formulated in combination with the network quality difference portrait and the accurate portrait of the user's family attributes. These strategies can not only take into account the user's network usage experience, but also fit their personal needs and preferences, thereby greatly improving user acceptance and satisfaction, and solving the problem that the existing broadband quality difference detection technology and user portrait construction method can only help operators identify and solve network problems to a certain extent, and cannot be effectively converted into a marketing strategy to improve user satisfaction and enhance competitiveness. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 A flowchart of a method for formulating a broadband marketing strategy according to Embodiment 1 of the present invention;

[0060] Figure 2 It is an overall flow chart of the method for formulating broadband marketing strategies according to an embodiment of the present invention;

[0061] Figure 3 A flowchart of broadband network quality analysis according to an embodiment of the present invention;

[0062] Figure 4 A flowchart of application matching according to an embodiment of the present invention;

[0063] Figure 5 A flowchart of user preference analysis according to an embodiment of the present invention;

[0064] Figure 6 A flowchart of household consumption capacity and resident population portrait according to an embodiment of the present invention;

[0065] Figure 7 A flowchart of judging potential household consumption concepts according to an embodiment of the present invention;

[0066] Figure 8 This is a flow chart of determining a household's potential consumption level according to an embodiment of the present invention;

[0067] Fig. 9 A flowchart of determining the size of a family population according to an embodiment of the present invention;

[0068] Fig.10 This is a flow chart of determining the attributes of a residential place according to an embodiment of the present invention;

[0069] Fig.11 A flowchart of child identification according to an embodiment of the present invention;

[0070] Fig.12 A flowchart of elderly identification according to an embodiment of the present invention;

[0071] Fig.13 A flowchart of active time analysis according to an embodiment of the present invention;

[0072] Fig.14 This is a schematic diagram of the structure of a device for formulating a broadband marketing strategy according to Embodiment 2 of the present invention;

[0073] Fig.15 This is a schematic diagram of the structure of a device for formulating a broadband marketing strategy according to Embodiment 3 of the present invention. DETAILED DESCRIPTION

[0074] In order to enable those skilled in the art to better understand the technical solution of the present invention, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0075] It should be understood that the specific embodiments and drawings described herein are only used to explain the present invention rather than to limit the present invention.

[0076] It can be understood that, in the absence of conflict, the various embodiments of the present invention and the various features in the embodiments can be combined with each other.

[0077] It can be understood that, for the convenience of description, the drawings of the present invention only show the parts related to the present invention, while the parts irrelevant to the present invention are not shown in the drawings.

[0078] It can be understood that each unit and module involved in the embodiments of the present invention may correspond to only one physical structure, or may be composed of multiple physical structures, or multiple units and modules may be integrated into one physical structure.

[0079] It can be understood that the terms "first", "second", etc. in the embodiments of the present invention are used to distinguish different objects, or to distinguish different processing of the same object, rather than to describe a specific order of objects.

[0080] It can be understood that, without conflict, the functions and steps marked in the flowcharts and block diagrams of the present invention may occur in an order different from that marked in the drawings.

[0081] It can be understood that the units and modules involved in the embodiments of the present invention can be implemented by software or hardware. For example, the units and modules can be located in a processor.

[0082] Embodiment 1:

[0083] This embodiment provides a method for formulating a broadband marketing strategy. Figure 1 As shown, the method includes:

[0084] Step S101: Obtaining EDPI (Elastic Deep Packet Inspect) data corresponding to target broadband account information within a preset time interval.

[0085] In this embodiment, the preset time interval can be one week, half a month, one month, etc., preferably one week. The EDPI collection equipment is located at the aggregation room node in each city. The equipment collects broadband data (ie, EDPI data) at the aggregation node close to the core network. As long as the traffic through broadband Internet access can be collected, it has the characteristics of wide coverage.

[0086] Step S102: obtaining the network quality score of the user in each time period within the preset time interval according to the EDPI data.

[0087] In this embodiment, the preset time interval is divided into multiple time periods, each time period can be one day, 12 hours, etc., preferably one day.

[0088] Optionally, obtaining the network quality score of the user in each time period within the preset time interval according to the EDPI data specifically includes:

[0089] According to the EDPI data of each time period corresponding to the target broadband account information, a network performance index of each time period is obtained, wherein the network performance index includes the number of ACK (ACKnowledge Character, confirmation character) packets, the number of packet losses, the first packet delay and the overall delay;

[0090] The network quality score of the user in each time period is calculated according to the number of ACK packets, the number of packet losses, the first packet delay and the overall delay in each time period.

[0091] In this embodiment, the first packet delay refers to the time interval from sending a request to receiving the first response data packet, and the overall delay refers to the time required to complete the entire data exchange.

[0092] Optionally, calculating the network quality score of the user in each time period according to the number of ACK packets, the number of packet losses, the first packet delay and the overall delay in each time period specifically includes:

[0093] Calculate the packet loss rate of each time period according to the number of ACK packets and the number of packet losses in each time period;

[0094] The network quality score for each time period is calculated according to the following formula:

[0095]

[0096] In the formula, Score PQ Indicates the network quality score, Score PLR Indicates the packet loss rate, ScoFe FBL Indicates the first packet delay, Score OL represents the overall delay, and α represents the correction factor.

[0097] In this embodiment, the packet loss rate can be calculated by the number of ACK packets and the number of lost packets, and the calculation formula is as follows:

[0098]

[0099] In this embodiment, by integrating the scores of the three key indicators of packet loss rate, first packet delay and overall delay, a comprehensive network quality score is obtained. This score not only reflects the objective performance of the network, but also takes into account the subjective feelings of the user. The network quality score formula is obtained by simulating and evaluating the user experience in different weak network scenarios. By matching the EDPI data of each test period with the user's subjective score, the setting value of the correction factor α in the network quality score formula can be obtained, which can then be used to adjust and optimize the accuracy of the network quality score, thereby better reflecting the user's real experience under specific network conditions.

[0100] Step S103: According to the network quality score of each time period, the network quality change of the user in the preset time interval is obtained to obtain the network quality profile of the user.

[0101] In this embodiment, taking the preset time interval as one week and the time period as one day as an example, the user's EDPI data for the past week can be obtained, and the user's daily network quality score can be calculated according to the aforementioned network quality score formula to obtain the user's network quality changes in the past week and obtain the user's network quality portrait.

[0102] Step S104: Perform user family attribute portrait analysis based on the EDPI data to obtain key dimensions for constructing an accurate user family attribute portrait, wherein the key dimensions include at least two of the user's Internet preferences, family consumption capacity and resident population, family member age distribution, and active time periods.

[0103] In this embodiment, through the detailed mining of EDPI data, it is possible to gain insight into key information such as the user's family's Internet habits, preferences, and active time periods, thereby outlining a three-dimensional network usage portrait for each family. The key dimensions used to construct an accurate portrait of the user's family attributes preferably include: user Internet preferences, family consumption capacity and resident population, family member age distribution, and active time periods.

[0104] Optionally, the user family attribute portrait analysis is performed based on the EDPI data to obtain key dimensions for constructing an accurate user family attribute portrait, including at least two of the following:

[0105] According to the application, access domain name and URL (Uniform Resource Locator) path information in the EDPI data, a standardized APP (Application) name and traffic type are obtained, and the usage time ratio of each type of APP is statistically analyzed through the standardized APP name and traffic type to obtain the user's Internet preference;

[0106] Determine the specific terminal devices and quantity used in the user's home according to the HTTP (Hypertext Transfer Protocol) header request information and the domain name in the EDPI data, and obtain the user's home consumption capacity and resident population based on the specific terminal devices and quantity used in the user's home;

[0107] Identify whether there are children and elderly people in the user's family based on the EDPI data, and obtain the age distribution of the user's family members;

[0108] The active time period of each family member in the user's home is analyzed according to the EDPI data, and a preset ARMA (Autoregressive Integrated Moving Average Model) model is used to predict the active time period of each family member at home in the future time interval.

[0109] In this embodiment, the standardized APP name refers to the APP name after being processed by certain rules and standards. These rules are usually to ensure the uniformity, readability and recognizability of the name. Specifically, the standardized APP name can be stored in the application information library in advance, and the standardized APP name can be obtained by matching the application in the EDPI data with the application information library, wherein the application in the EDPI data reflects the APP name currently being used by the user, and the domain name and URL path in the EDPI data are regularized and parsed to obtain the APP name and the partition path in the APP (such as watching videos, watching live broadcasts, etc.), which can be used as a supplement to the application field to obtain the corresponding traffic type. The traffic type refers to the different types of data traffic or behavior patterns generated by the user in the APP, which can include video traffic, live broadcast traffic, browsing traffic, etc. Through the standardized APP name and traffic type, the proportion of the use time of each type of APP is statistically analyzed. According to the distribution of the use time proportion, through the pre-set preference judgment algorithm, we can dig out the user's Internet preferences.

[0110] Optionally, determining the specific terminal devices and quantity used in the user's home according to the HTTP header request information and the domain name in the EDPI data specifically includes:

[0111] Extracting a terminal identifier from the HTTP header request information and the domain name in the EDPI data;

[0112] Compare the extracted terminal identifier with the pre-established terminal information database to obtain the specific terminal devices and quantity used in the user's home;

[0113] The obtaining of the household consumption capacity and resident population of the user based on the specific terminal devices and the number of terminal devices used in the household of the user specifically includes:

[0114] Calculate a comprehensive score of the potential consumption concept of the user's family based on the specific terminal devices and quantity used in the user's family;

[0115] The potential consumption concept of the user's family is obtained by comprehensively scoring the potential consumption concept of the user's family;

[0116] Calculate the potential consumption level score of the user's family based on the potential consumption concept of the user's family and the broadband package information of the user's family;

[0117] The potential consumption level of the user's family is obtained according to the potential consumption level score of the user's family;

[0118] Identify the size of the permanent population of the user's home based on the active time period of each terminal device among the specific terminal devices used in the user's home;

[0119] According to the broadband installation address of the user's family, the residential property of the user's family is obtained with the help of a real estate information platform, and the residential consumption level of the user's family is obtained based on the residential property.

[0120] In this embodiment, regular expression technology is used to extract terminal identifiers from HTTP header request information and domain names in EDPI data. These identifiers usually include key information such as device model and operating system version, which are important clues for identifying the type of user terminal. The extracted terminal identifier is compared with a pre-established terminal information library. The terminal information library covers detailed data of various types of devices such as mainstream smartphones, tablets, notebooks, smart TVs, game consoles, etc. on the market. By comparing, the specific terminal devices and quantities used in the user's home can be determined.

[0121] Optionally, the comprehensive score of the potential consumption concept of the user's family is calculated based on the specific terminal devices and the number of terminal devices used in the user's family, and is specifically calculated according to the following formula:

[0122]

[0123] In the formula, C represents the comprehensive score of the user's family's potential consumption concept, n represents the number of specific terminal devices used in the user's family, and w i represents the weight of the i-th terminal device, C i Represents the household consumption concept reflected by the i-th terminal device.

[0124] In this embodiment, w i Determined by the market position of the terminal device, for example, the high-end can be 3, the mid-end can be 2, and the low-end can be 1. i The calculation formula can be as follows:

[0125]

[0126] Where D i is the depreciation rate of the i-th terminal equipment, and the formula is Reflects the degree of reduction in the value of terminal equipment over time; T i is the time from the launch of the i-th terminal device to the present, in months, and the formula is T = current time - launch time; T maxis the longest life of the terminal equipment, which is used to normalize to the interval [0, 1]. The default value is usually 5 years, i.e. 60 months. D and w T The weight is adjustable.

[0127] In this embodiment, the potential consumption concept of the user's family can be obtained through the score range corresponding to the comprehensive score of the family consumption concept. For example, the interval [0, 40) can be considered as frugal, [40, 70) is ordinary, and [70, 100] is trendy.

[0128] Optionally, the potential consumption level score of the user's family is calculated based on the potential consumption concept of the user's family and the broadband package information of the user's family, and is specifically calculated according to the following formula:

[0129] L=w C ×C+w p ×P

[0130] Where L represents the potential consumption level score of the user's family, w C represents the weight of the comprehensive score of potential consumption concept, P represents the score of the broadband package information of the user's family, and w p Indicates the weight of the broadband package information score.

[0131] In this embodiment, the broadband package information score P of the user's family can be calculated according to the following formula:

[0132] P=α×S p +β×S s +γ×S d

[0133] In the formula, S p Indicates the monthly fee of the broadband package, S s Indicates the number of additional services in the broadband package, S d represents the package period of the broadband package, α, β, and γ represent S p , S s , S d The weight it occupies.

[0134] In this embodiment, the potential consumption level of the user's family can be obtained through the score interval corresponding to the potential consumption level score. For example, the interval [0, 30] is a low level, [30, 70] is a medium level, and (70, 100] is a high level.

[0135] Optionally, the identifying the size of the permanent population of the user's home based on the active time period of each terminal device among the specific terminal devices used in the user's home specifically includes:

[0136] Taking terminal devices as units, count the active time periods of each terminal device in the user's home, and calculate the average usage time ratio of each terminal device within the preset time intervals of various APP applications;

[0137] According to the active time period of each terminal device and the average usage time ratio of each terminal device within the preset time interval of each type of APP application, the usage patterns of the terminal devices are clustered to obtain the permanent population size of the user's family.

[0138] In this embodiment, the active time period of each terminal device is counted in units of terminal devices, and mapped to multiple time periods in a day to obtain a time period dimension vector. The average usage time ratio of each terminal device in various APP applications in a week is counted to obtain an application dimension vector. After splicing the time period dimension vector and the application dimension vector of a week, the terminal vector is obtained. The cosine distance and DBSCAN algorithm are used to cluster the usage patterns of the terminal devices, and each category represents a family member. From this, the size of the resident population of the family can be roughly determined: single-person family, 2-person family, small-scale family of 3, medium-sized family of 4-5 people, large-scale family of 6-7 people, super-large family of more than 7 people, etc.

[0139] In this embodiment, the home broadband installation address can be used to further explore the detailed information of the living environment. With the help of real estate information platforms such as Beike House Search and Anjuke, we can automatically capture the geographical location, rent, house price and other data related to the address. Based on this information, the residential property can be subdivided into urban village rental houses, self-built houses, ordinary community houses, mid-range community houses, high-end community houses, etc. Through this information, the residential consumption level of the user's family can be obtained.

[0140] In this embodiment, the same method as before is used to identify applications and terminal devices, and then it is determined whether there are parenting and learning guidance apps in the identified apps, or whether there are children's watches in the identified terminals, so as to determine whether there are children in the family. Similarly, the same method as before is used to identify terminal devices, and then in the identified terminal devices, after matching with the terminal information database, it is determined whether there are low-end mobile phones. If so, it is continued to determine whether there is a TV and whether there is TV traffic during working hours, so as to determine whether there are elderly people in the family.

[0141] In this embodiment, the active time period of each family member is tracked based on EDPI data. Taking one week's data as an example, by performing statistical analysis on the timestamp data at an hourly granularity, we can capture the user's daily active time period, and then infer their living habits, such as getting up early or staying up late. This can provide a basis for optimizing push strategies and improving information acceptance. On this basis, ARMA and other time series algorithms are used to predict the daily active time of the member in the next week. Focus on the distribution of active time periods at home by the owner (generally set as non-elderly children, people who have a lot of say in the family) to facilitate marketing strategy formulation.

[0142] Step S105: constructing an accurate portrait of the user's family attributes according to the key dimensions used to construct an accurate portrait of the user's family attributes;

[0143] Step S106: Formulate a broadband marketing strategy based on the network quality profile and the accurate profile of the user's family attributes.

[0144] In this embodiment, in order to fully utilize the synergistic effect of the network quality difference portrait and the accurate portrait of the user's family attributes and improve the accuracy and effectiveness of the marketing strategy, a broadband marketing strategy is formulated based on the network quality difference portrait and the accurate portrait of the user's family attributes. These strategies can not only take into account the user's network usage experience, but also fit their personal needs and preferences, thereby greatly improving the user's acceptance and satisfaction.

[0145] It should be noted that during the research and practice of the prior art, the inventors found that the prior art had the following pain points:

[0146] 1. Existing technologies often treat broadband quality detection and user profiling as two independent research fields, and few studies combine the two to innovate marketing strategies. This separation makes it impossible to fully utilize the synergy between the two, limiting the accuracy and effectiveness of marketing strategies.

[0147] Specifically, in practice, the inventor selected two communities and tried one of the solutions respectively, and the outbound call conversion rates were low. Community A only used broadband quality detection data for marketing. Although it could identify the target user group more accurately, due to the lack of understanding of the user's family portrait, the marketing plan was misaligned with the user's actual situation and needs. As a result, even for users who are willing to upgrade, the marketing may fail because the solution does not meet their needs, and the final conversion rate is only 14%; Community B only uses user portrait data for marketing, and can formulate a plan that users are more satisfied with, but many users are satisfied with the current network quality and believe that there is no need to upgrade, so they refuse the marketing promotion, and the final conversion rate is only 11%.

[0148] Therefore, the inventor believes that broadband quality detection and user portrait drawing should cooperate with each other to form a synergistic effect. A single solution cannot fully capture the characteristics and needs of users, which is the main reason for the low conversion rate of the two marketing activities.

[0149] 2. In terms of user profiling, there are problems such as insufficient data coverage (for example, solutions that rely on WIFI data, some broadband users may not use WIFI), multi-source data alignment and consistency issues, and manual determination of feature importance and thresholds that may result in inaccurate results.

[0150] In order to solve the above pain points, the purpose of the present invention is to:

[0151] 1. Combine broadband quality detection and user profiling to output more accurate user features

[0152] The present invention is based on the data of EDPI flexible packet detection technology. The data is collected at the exit of the metropolitan area network. On the one hand, broadband detection can be performed through network indicators such as EDPI's first packet delay and number of packet losses. On the other hand, the user's terminal, Internet preference, active time and other information can be analyzed through EDPI's HTTP request header information, domain name, timestamp, application and other information to obtain a user portrait.

[0153] 2. Solve the problems of limited data coverage, multi-source data alignment and consistency

[0154] The present invention is based only on EDPI data, thus solving the problem of multi-source data alignment and consistency; at the same time, as long as the traffic through broadband Internet access can be collected, the problem of limited coverage is solved.

[0155] Based on the above purpose, the present invention simultaneously detects broadband network quality and obtains user portraits by analyzing EDPI data, and has better practical application effects. In the process of continuous iterative optimization, the inventors conducted field tests and application practices in multiple communities. The practical results show that this solution can provide a more comprehensive user description, thanks to the dual advantages of integrating network quality detection and user family portrait analysis. Through the careful mining of EDPI data, we can not only evaluate the quality of network use, but also gain insight into key information such as user family Internet habits, preferences, and active time periods, thereby outlining a three-dimensional network usage portrait for each family. Thanks to this comprehensive and in-depth understanding of users, we can formulate more accurate and personalized marketing strategies. These strategies not only take into account the user's network usage experience, but also fit their personal needs and preferences, greatly improving user acceptance and satisfaction. In the latest marketing campaign, the conversion rate reached 27%, which far exceeds the results of applying broadband quality detection or user portrait solutions alone. This achievement not only provides new marketing ideas for broadband operators, but also opens up new paths for service optimization and product innovation in the communications industry.

[0156] In a specific embodiment, by deeply analyzing the different fields of EDPI information, the broadband network quality score and the accurate portrait of the user's family attributes are obtained at the same time, laying a solid data foundation for providing personalized services and optimizing marketing strategies. The EDPI collection equipment is located at the nodes of the aggregation room in each city. The overall process of the broadband marketing strategy formulation method is as follows: Figure 2 As shown, the following steps may be included:

[0157] S1, broadband network quality analysis:

[0158] like Figure 3 As shown in the figure, we adopted a systematic approach to simulate and evaluate the user experience in different weak network scenarios, which is divided into several key steps:

[0159] 1. Test environment setup: We used an optical modem and a router with OpenWRT to build a simple test environment in a controlled laboratory environment, and used the same mobile phone for testing. Here, we used the TC (Traffic Control) tool, which is a common open source tool widely used in network research. It can finely control network parameters, including setting various delays and packet loss rates. By adjusting these parameters, we can simulate a variety of weak network scenarios, from slight network fluctuations to severe packet loss, covering various network conditions that users may encounter.

[0160] 2. Obtain subjective scores: We carefully observe and record each simulated weak network scenario to obtain subjective quality scores. By using different apps in the same weak network environment, such as Bilibili, Douyin, Taobao, JD.com, WeChat, QQ and other commonly used apps, we record key information such as freeze frequency, freeze duration, loading time, and give a subjective score based on this information. This score reflects the intuitive changes in user experience under different network environments.

[0161] 3. Quantify the impact of network indicators on experience: We correspond the EDPI data of each test period with the subjective score. Among them, we extract four core indicators: the number of ACK packets, the number of packet losses, the first packet delay and the overall delay. The first packet delay refers to the time interval from sending a request to receiving the first response data packet, and the overall delay is the time required to complete the entire data exchange. The packet loss rate can be calculated by the number of ACK packets and the number of packet losses. We divide these network performance indicators into different intervals and assign corresponding scores to each interval to quantify the impact of different delays and packet loss levels on user experience. The packet loss rate is calculated as follows:

[0162]

[0163] 4. Calculate the network quality score: Integrate the scores of the three key indicators, packet loss rate, first packet delay and overall delay, to get the broadband quality score. This score not only reflects the objective performance of the network, but also takes into account the subjective feelings of users. The calculation method is as follows:

[0164]

[0165] Among them, α is the correction factor.

[0166] 5. Obtain quality changes: In actual application, we obtain the user's data from the past week, and calculate the user's daily quality score based on the aforementioned formula, obtain the user's network quality changes from the past week, and obtain the user's network quality portrait, so that marketers can formulate targeted strategies.

[0167] S2, User Family Attribute Profile Analysis:

[0168] 1. Internet Preference Mining

[0169] (1) Application matching: Combining the application, access domain name and URL path information in the EDPI data, we have built a multi-dimensional data analysis framework. First, the application field reflects the name of the APP that the user is currently using; then, by regularizing the domain name and URL path, we obtain the APP name and the partition path within the APP (such as watching videos, watching live broadcasts, etc.) as a supplement to the application field. Finally, match it with the application information library to obtain the standardized APP name and traffic type. The specific flow chart is as follows. Figure 4 shown.

[0170] (2) User preference modeling: By standardizing the APP name and traffic type, we statistically analyze the proportion of APP usage time. Based on the distribution of usage time, we can mine the user's Internet preferences through the preference judgment algorithm we set. The flowchart of user preference analysis is shown in the figure below. Figure 5 As shown. These preference patterns can provide support for personalized recommendations and precision marketing. Preferences include but are not limited to game preferences, live broadcast preferences, video preferences, online shopping preferences, travel preferences, literary preferences, etc., and can also output the top 3 commonly used apps, or app preferences or top 3 apps in different time periods. Considering that the user is likely not a single preference user, we use the difference threshold and activation threshold to assist in the judgment to make it more in line with the actual situation. If the APP category with the largest and second largest duration is greater than the difference threshold, the user can be considered a heavy single preference user; otherwise, it may be a multi-preference user, in which case the output is based on the duration ratio being greater than the activation threshold.

[0171] The preference judgment logic is as follows:

[0172] Algorithm: User preference judgment algorithm

[0173] Input: List of usage time of various APPs

[0174] S=[(APP1,T1),(APP2,T2),…,(APP n , T n )], difference threshold τ1, activation threshold T2 Output: User preference list L

[0175] pseudocode:

[0176] 1) Determine the category preference type N

[0177] 2) For each preference category N i , find out the total duration of this type of APP and calculate its proportion P i

[0178] 3) Get the APP category N with the largest time usage max Its duration ratio P max, and the second largest category of APPs in terms of time usage N second Its duration ratio P second

[0179]

[0180] 2. Household consumption capacity and resident population profile

[0181] The flow chart of household consumption capacity and resident population portrait is as follows Figure 6 As shown, including:

[0182] (1) Data source analysis: First, we use regular expression technology to extract terminal identifiers from the HTTP header request information and domain names in the EDPI data. These identifiers usually include key information such as device model and operating system version, which are important clues for identifying user terminal types.

[0183] (2) Terminal information database comparison: We then compare the extracted terminal identifier with our pre-established terminal information database. This terminal information database covers detailed data on mainstream smartphones, tablets, notebooks, smart TVs, game consoles, and other devices on the market. Through the comparison, we can determine the specific terminal devices and quantity used in the user's home.

[0184] (3) Family attribute analysis: Based on the type and number of terminal devices, we can further infer the following family attribute profiles:

[0185] a) Potential consumer concepts

[0186] The flowchart of judging the potential consumption concept of a family is as follows Figure 7 As shown, specifically, combined with our terminal database, we can further obtain the launch time, launch price, current market price, and market positioning (high-end, mid-end, and low-end) of the identified terminal. Combining these four indicators, we can determine the potential consumption concept of the user's family (thrifty, trendy, and ordinary):

[0187]

[0188] in:

[0189] C is the comprehensive score of the family's consumption concept. According to the mapping of the score range, the user's current consumption concept can be approximately obtained. For example, the interval [0, 40) can be considered as frugal, [40, 70) is ordinary, and [70, 100] is trendy. n is the number of terminals owned by the family.

[0190] w i It represents the weight of the ith terminal, which is determined by the market position of the terminal device and has the following values:

[0191]

[0192] C i It represents the household consumption concept reflected by the i-th terminal, and the calculation formula is as follows:

[0193]

[0194] in:

[0195] D i is the depreciation rate of the i-th terminal equipment, and the formula is It reflects the extent of the terminal equipment's value reduction over time.

[0196] T i is the time from the launch of the i-th terminal device to the present, in months, and the formula is T = current time - launch time;

[0197] T max It is the maximum lifespan of the terminal equipment and is used to normalize it to the interval [0, 1]. The default value is usually 5 years, or 60 months.

[0198] w D and w T The weight is adjustable.

[0199] b) Potential consumption level

[0200] The flowchart for judging the potential consumption level of a family is as follows: Figure 8 Specifically, after obtaining the family's potential consumption concept in the first step, we can further predict the family's potential consumption level by combining the family package information we have and using a weighted scoring method:

[0201] L=w C ×C+w p ×P

[0202] in:

[0203] L is the consumption level score, the interval [0, 30] is a low level, [30, 70] is a medium level, and (70, 100] is a high level.

[0204] C is the consumption concept score;

[0205] P is the broadband package information score, which is calculated based on the monthly fee, number of additional services, duration, etc. The calculation method is as follows:

[0206] P=α×S p +β×S s +γ×S d

[0207] in:

[0208] S p It indicates the monthly fee of the package, which is the most intuitive indicator;

[0209] S s Indicates the number of additional services (such as TV fees, cameras, home primary and secondary cards, etc.). The more the number, the more user needs and the higher the corresponding consumption level;

[0210] S d Indicates the package duration, calculated by subtracting the current time from the contract expiration date. The shorter the term, the higher the user's willingness to consume.

[0211] c) Population size

[0212] The flow chart for determining family population size is as follows Fig. 9 As shown, specifically, through the number of terminals, active time periods, and applications identified in the first step, based on a week's data and combined with daily life experience (basically those who are at home during the day are the elderly, preschool children or full-time housewives, and adults have office software on their phones, etc.), we can roughly infer the population size of the family.

[0213] First, we count the active time periods of each terminal and map them to four time periods per day (7am to 1pm, 1pm to 7pm, 7pm to 1am, and 1am to 7am) to get a 4-dimensional vector. For example, if the terminal is only active from 7pm to 1am, the vector is [0, 0, 1, 0].

[0214] Then, combined with the algorithm of application identification in the first step, the average usage time ratio of each type of application on each terminal in a week is counted to obtain a vector of no more than 20 dimensions. For example, if we only focus on game, social, office, and video apps, and the average daily usage time of game apps on a terminal in a week is 60 minutes, the usage time of social apps is 240 minutes, the usage time of office apps is 240 minutes, and the usage time of video apps is 60 minutes, then a 4-dimensional vector of [0.1, 0.4, 0.4, 0.1] will be obtained, and the number of dimensions is equal to the number of application types we focus on.

[0215] After concatenating the time period dimension vector and the application dimension vector, we get the terminal vector. We use the cosine distance and DBSCAN algorithm to cluster the terminal usage patterns, and each category represents a family member. From this, we can roughly determine the population size of the family: single-person family, 2-person family, small-scale 3-person family, medium-sized 4-5-person family, large-scale 6-7-person family, and super-large family with more than 7 people.

[0216] Cosine distance formula:

[0217]

[0218] Among them, x, y are vectors in n-dimensional space, namely the terminal vectors mentioned above.

[0219] The pseudo code of DBSCAN clustering algorithm is as follows:

[0220] Algorithm: DBSCAN clustering algorithm

[0221] Input: A dataset D containing n terminal vectors, a radius parameter ε, and a cluster density threshold MinPts

[0222] Output: A collection of density-based clusters representing household populations

[0223] 1) Mark all vectors as unvisited;

[0224] 2) Randomly select an unvisited vector p and mark it as visited

[0225] 3)If there are at least MinPts vectors in the ε-range of p

[0226] 4) Create a new cluster C and add p to C

[0227] 5) Let N be the set of vectors within the ε-range of p

[0228] 6)For each vector p′ in N

[0229]

[0230] 7) Else mark p as noise

[0231] 8) Repeat steps 2)-7) until all vectors are marked as visited and the algorithm ends.

[0232] d) Residential location attributes

[0233] The flowchart of residential place attribute judgment is as follows Fig.10 As shown, specifically, using the home broadband installation address, we can further explore the detailed information of the living environment. With the help of real estate information platforms such as Beike, Anjuke, etc., we can automatically capture the geographic location, rent, house price and other data related to the address. Based on this information, the attributes of the residential place can be subdivided into urban village rental houses, self-built houses, ordinary community houses, mid-range community houses, high-end community houses, etc. Through this information, we can also evaluate the family's residential consumption level from another perspective. The judgment logic is as follows:

[0234]

[0235] 3. Age distribution of family members

[0236] (1) Child identification: Fig.11 As shown, the same method as before is used to identify applications and terminals, and then determine whether there are parenting or study guidance apps in the identified apps, or determine whether there are children's watches in the identified terminals, etc., to determine whether there are children in the family.

[0237] (2) Elderly identification: Fig.12 As shown, the same method as before is used to identify the terminal, and then among the identified terminals, after matching with the terminal information database, it is determined whether there is a low-end mobile phone. If so, it is further determined whether there is a TV and whether there is TV traffic during working hours, so as to determine whether there are elderly people in the family.

[0238] 4. Active time analysis

[0239] The flow chart of active time analysis is as follows Fig.13 As shown, including:

[0240] (1) Identification of active time periods: Based on the assumption that the background traffic of home devices is evenly distributed, combined with the algorithm of the family population size in the second part, the active time periods of each family member are tracked. Taking one week's data as an example, by statistically analyzing the timestamp data at the hourly granularity, we can capture the user's daily active time period and then infer their living habits, such as getting up early or staying up late. This can provide a basis for optimizing push strategies and improving information acceptance.

[0241] (2) Active time period prediction: Based on the first step, the ARMA and other time series algorithms are used to predict the daily active time of the member in the next week. The focus is on the distribution of active time periods of the owner (generally set as non-elderly children and people with more say in the family) at home to facilitate the formulation of marketing strategies.

[0242] The ARMA model, or autoregressive moving average model, is a commonly used time series analysis algorithm used to predict future data. Its general form can be expressed as:

[0243]

[0244] in:

[0245] yt is the observed value at time point t, which is also the predicted value;

[0246] c is a constant term, usually set to 0;

[0247] φ is the autoregressive coefficient and θ is the moving average coefficient

[0248] ε is a white noise error term, which is usually assumed to be an independent and identically distributed random variable with mean 0 and variance 1.

[0249] It should be noted that this technical solution deeply analyzes EDPI data to obtain changes in user home network quality and accurate family portraits. In the family portrait, it can not only accurately identify the terminal devices and consumption attributes of the user's home, but also gain insight into the user's online behavior patterns and interest preferences, providing operators with a set of comprehensive user portrait construction tools, which helps to improve the level of service personalization and the accuracy of marketing activities.

[0250] The method for formulating a broadband marketing strategy provided by an embodiment of the present invention first obtains the flexible packet detection technology EDPI data corresponding to the target broadband account information within a preset time interval; then obtains the network quality difference score of the user in each time period within the preset time interval based on the EDPI data; and obtains the network quality change of the user within the preset time interval based on the network quality difference score of each time period, and obtains the network quality difference portrait of the user; at the same time, based on the EDPI data, the user's family attribute portrait analysis is performed to obtain key dimensions for constructing an accurate portrait of the user's family attributes, and the key dimensions include at least two of the user's Internet preferences, family consumption capacity and resident population, age distribution of family members, and active time periods; and constructs an accurate portrait of the user's family attributes based on the key dimensions for constructing an accurate portrait of the user's family attributes; finally, a broadband marketing strategy is formulated based on the network quality difference portrait and the accurate portrait of the user's family attributes. The present invention can simultaneously obtain a network quality difference portrait and an accurate portrait of the user's family attributes through the analysis of EDPI data. Since the combination of the two can provide more comprehensive user characteristics, more accurate and personalized marketing strategies can be formulated in combination with the network quality difference portrait and the accurate portrait of the user's family attributes. These strategies can not only take into account the user's network usage experience, but also fit their personal needs and preferences, thereby greatly improving user acceptance and satisfaction, and solving the problem that the existing broadband quality difference detection technology and user portrait construction method can only help operators identify and solve network problems to a certain extent, and cannot be effectively converted into a marketing strategy to improve user satisfaction and enhance competitiveness.

[0251] Embodiment 2:

[0252] like Fig.14 As shown, this embodiment provides a broadband marketing strategy formulation device, which is used to execute the broadband marketing strategy formulation method, including:

[0253] A data acquisition module 11 is used to acquire flexible packet detection technology EDPI data corresponding to target broadband account information within a preset time interval;

[0254] A quality difference score acquisition module 12, connected to the data acquisition module 11, is used to acquire the network quality difference score of the user in each time period within the preset time interval according to the EDPI data;

[0255] The quality difference portrait acquisition module 13 is connected to the quality difference score acquisition module 12, and is used to obtain the network quality change of the user in the preset time interval according to the network quality difference score of each time period, and obtain the network quality difference portrait of the user;

[0256] A key dimension acquisition module 14, connected to the data acquisition module 11, is used to perform user family attribute portrait analysis based on the EDPI data to obtain key dimensions for constructing an accurate user family attribute portrait, wherein the key dimensions include at least two of the user's Internet preferences, family consumption capacity and resident population, family member age distribution, and active time period;

[0257] A family portrait construction module 15, connected to the key dimension acquisition module 14, is used to construct an accurate portrait of the user's family attributes according to the key dimensions used to construct an accurate portrait of the user's family attributes;

[0258] The marketing strategy formulation module 16 is connected to the quality difference portrait acquisition module 13 and the family portrait construction module 15, and is used to formulate a broadband marketing strategy based on the network quality difference portrait and the accurate portrait of the user's family attributes.

[0259] Optionally, the quality difference score acquisition module 12 includes:

[0260] A network performance indicator acquisition unit, configured to obtain a network performance indicator for each time period according to the EDPI data of each time period corresponding to the target broadband account information, wherein the network performance indicator includes the number of ACK packets, the number of packet losses, the first packet delay and the overall delay;

[0261] The quality difference score calculation unit is used to calculate the network quality difference score of the user in each time period according to the number of ACK packets, the number of packet losses, the first packet delay and the overall delay in each time period.

[0262] Optionally, the quality difference score calculation unit specifically includes:

[0263] A packet loss rate calculation unit, configured to calculate the packet loss rate of each time period according to the number of ACK packets and the number of lost packets in each time period;

[0264] The first score calculation unit is used to calculate the network quality score of each time period according to the following formula:

[0265]

[0266] In the formula, Score PQ Indicates the network quality score, Score PLR Indicates the packet loss rate, Score FBL Indicates the first packet delay, Score OL represents the overall delay, and α represents the correction factor.

[0267] Optionally, the key dimension acquisition module 14 includes at least two of the following:

[0268] An online preference acquisition unit is used to obtain a standardized application program APP name and traffic type according to the application, access domain name and uniform resource locator URL path information in the EDPI data, and statistically analyze the proportion of usage time of various types of APPs through the standardized APP name and traffic type to obtain the user's online preference;

[0269] A household attribute acquisition unit, configured to determine the specific terminal devices and the number of terminal devices used in the household of the user according to the Hypertext Transfer Protocol (HTTP) header request information and the domain name in the EDPI data, and obtain the household consumption capacity and the resident population of the user based on the specific terminal devices and the number of terminal devices used in the household of the user;

[0270] An age distribution acquisition unit, used to identify whether there are children and elderly people in the user's family according to the EDPI data, and obtain the age distribution of the user's family members;

[0271] The active time period acquisition unit is used to analyze the active time period of each family member in the user's family according to the EDPI data, and use a preset autoregressive moving average model ARMA model to predict the active time period of each family member at home in the future time interval.

[0272] Optionally, the family attribute acquisition unit includes:

[0273] An identifier extraction unit, used to extract a terminal identifier from the HTTP header request information and the domain name in the EDPI data;

[0274] The home terminal acquisition unit is used to compare the extracted terminal identifier with the pre-established terminal information database to obtain the specific terminal devices and quantities used in the user's home:

[0275] A consumption concept comprehensive scoring unit, used to calculate a comprehensive score of potential consumption concepts of the user's family based on the specific terminal devices and the number of terminal devices used in the user's family;

[0276] A potential consumption concept acquisition unit, used to obtain the potential consumption concept of the user's family according to the comprehensive score of the potential consumption concept of the user's family;

[0277] A potential consumption level scoring unit, used to calculate the potential consumption level score of the user's family based on the potential consumption concept of the user's family and the broadband package information of the user's family;

[0278] A potential consumption level acquisition unit, used to obtain the potential consumption level of the user's family according to the potential consumption level score of the user's family;

[0279] A population size acquisition unit, configured to identify the resident population size of the user's home based on the active time period of each terminal device among the specific terminal devices used in the user's home;

[0280] The residential consumption level acquisition unit is used to acquire the residential property of the user's family according to the broadband installation address of the user's family with the help of a real estate information platform, and obtain the residential consumption level of the user's family based on the residential property.

[0281] Optionally, the consumption concept comprehensive scoring unit calculates the potential consumption concept comprehensive score of the user's family according to the following formula:

[0282]

[0283] In the formula, C represents the comprehensive score of the user's family's potential consumption concept, n represents the number of specific terminal devices used in the user's family, and w i represents the weight of the i-th terminal device, C i represents the household consumption concept reflected by the i-th terminal device;

[0284] The potential consumption level scoring unit specifically calculates the potential consumption level score of the user's family according to the following formula:

[0285] L=w C ×C+w p ×P

[0286] Where L represents the potential consumption level score of the user's family, w C represents the weight of the comprehensive score of potential consumption concept, P represents the score of the broadband package information of the user's family, and w p Indicates the weight of the broadband package information score.

[0287] Optionally, the broadband package information score P of the user's family is calculated according to the following formula:

[0288] P=α×S p +β×S s +γ×S d

[0289] In the formula, S p Indicates the monthly fee of the broadband package, S sIndicates the number of additional services in the broadband package, S d represents the package period of the broadband package, α, β, and γ represent S p , S s , S d The weight it occupies.

[0290] Optionally, the population size acquisition unit includes:

[0291] A statistical unit, used to count the active time periods of each terminal device in the user's home, and calculate the average usage time ratio of each terminal device within the preset time interval of each type of APP application;

[0292] The clustering unit is used to cluster the usage patterns of the terminal devices according to the active time period of each terminal device and the average usage time ratio of each terminal device within the preset time interval of various APP applications, so as to obtain the permanent population size of the user's family.

[0293] Embodiment 3:

[0294] refer to Fig.15 This embodiment provides a broadband marketing strategy formulation device, including a memory 21 and a processor 22, the memory 21 stores a computer program, and the processor 22 is configured to run the computer program to execute the broadband marketing strategy formulation method in Example 1.

[0295] The memory 21 is connected to the processor 22. The memory 21 may be a flash memory, a read-only memory or other memory. The processor 22 may be a central processing unit or a single-chip microcomputer.

[0296] Embodiment 4:

[0297] This embodiment provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the method for formulating a broadband marketing strategy in the above-mentioned embodiment 1 is implemented.

[0298] The computer-readable storage medium includes volatile or non-volatile, removable or non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, computer program modules or other data). Computer-readable storage media include, but are not limited to, RAM (Random Access Memory), ROM (Read-Only Memory), EEPROM (Electrically Erasable Programmable read only memory), flash memory or other memory technology, CD-ROM (Compact Disc Read-Only Memory), digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer.

[0299] In summary, the broadband marketing strategy formulation method, device and readable storage medium provided by the embodiments of the present invention first obtain the flexible packet detection technology EDPI data corresponding to the target broadband account information within a preset time interval; then obtain the network quality difference score of the user in each time period within the preset time interval based on the EDPI data; and obtain the network quality change of the user in the preset time interval based on the network quality difference score of each time period, and obtain the user's network quality difference portrait; at the same time, based on the EDPI data, perform user family attribute portrait analysis to obtain key dimensions for constructing an accurate portrait of the user's family attributes, the key dimensions include at least two of the user's Internet preferences, family consumption capacity and resident population, age distribution of family members, and active time periods; and construct an accurate portrait of the user's family attributes based on the key dimensions for constructing an accurate portrait of the user's family attributes; finally, formulate a broadband marketing strategy based on the network quality difference portrait and the accurate portrait of the user's family attributes. The present invention can simultaneously obtain a network quality difference portrait and an accurate portrait of the user's family attributes through the analysis of EDPI data. Since the combination of the two can provide more comprehensive user characteristics, more accurate and personalized marketing strategies can be formulated in combination with the network quality difference portrait and the accurate portrait of the user's family attributes. These strategies can not only take into account the user's network usage experience, but also fit their personal needs and preferences, thereby greatly improving user acceptance and satisfaction, and solving the problem that the existing broadband quality difference detection technology and user portrait construction method can only help operators identify and solve network problems to a certain extent, and cannot be effectively converted into a marketing strategy to improve user satisfaction and enhance competitiveness.

[0300] It is to be understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present invention, but the present invention is not limited thereto. For those of ordinary skill in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A method for formulating a broadband marketing strategy, characterized in that: The method comprises: Obtaining flexible packet inspection technology EDPI data corresponding to target broadband account information within a preset time interval; Obtaining a network quality score of the user in each time period within the preset time interval according to the EDPI data; According to the network quality score of each time period, the network quality change of the user in the preset time interval is obtained to obtain the network quality profile of the user; Performing a user family attribute profile analysis based on the EDPI data to obtain key dimensions for constructing an accurate user family attribute profile, wherein the key dimensions include at least two of the user's Internet preferences, family consumption capacity and resident population, family member age distribution, and active time period; Constructing an accurate portrait of the user's family attributes according to the key dimensions used to construct an accurate portrait of the user's family attributes; and, A broadband marketing strategy is formulated based on the network quality profile and the accurate profile of the user's family attributes.

2. The method according to claim 1, characterized in that The obtaining, according to the EDPI data, a network quality score of the user in each time period within the preset time interval specifically includes: According to the EDPI data of each time period corresponding to the target broadband account information, the network performance index of each time period is obtained, wherein the network performance index includes the number of ACK packets, the number of packet losses, the first packet delay and the overall delay; The network quality score of the user in each time period is calculated according to the number of ACK packets, the number of packet losses, the first packet delay and the overall delay in each time period.

3. The method according to claim 2, characterized in that The calculating of the network quality score of the user in each time period according to the number of ACK packets, the number of packet losses, the first packet delay and the overall delay in each time period specifically includes: Calculate the packet loss rate of each time period according to the number of ACK packets and the number of packet losses in each time period; The network quality score for each time period is calculated according to the following formula: In the formula, Score PQ Indicates the network quality score, Score PLR Indicates the packet loss rate, Score FBL Indicates the first packet delay, Score OL represents the overall delay, and α represents the correction factor.

4. The method according to claim 1, characterized in that: The user family attribute portrait analysis based on the EDPI data is performed to obtain key dimensions for constructing an accurate user family attribute portrait, including at least two of the following: Obtaining a standardized application program APP name and traffic type based on the application, access domain name and uniform resource locator URL path information in the EDPI data, and statistically analyzing the usage time ratio of each type of APP through the standardized APP name and traffic type to obtain the user's Internet preferences; Determine the specific terminal devices and quantity used in the user's home according to the Hypertext Transfer Protocol HTTP header request information and the domain name in the EDPI data, and obtain the user's home consumption capacity and resident population based on the specific terminal devices and quantity used in the user's home; Identify whether there are children and elderly people in the user's family based on the EDPI data, and obtain the age distribution of the user's family members; The active time period of each family member in the user's home is analyzed according to the EDPI data, and the preset autoregressive moving average model ARMA model is used to predict the active time period of each family member at home in the future time interval.

5. The method according to claim 4, characterized in that The determining of the specific terminal devices and the number of the terminal devices used in the user's home according to the Hypertext Transfer Protocol HTTP header request information and the domain name in the EDPI data specifically includes: Extracting a terminal identifier from the HTTP header request information and the domain name in the EDPI data; Compare the extracted terminal identifier with the pre-established terminal information database to obtain the specific terminal devices and quantity used in the user's home; The obtaining of the household consumption capacity and resident population of the user based on the specific terminal devices and the number of terminal devices used in the household of the user specifically includes: Calculate a comprehensive score of the potential consumption concept of the user's family based on the specific terminal devices and quantity used in the user's family; The potential consumption concept of the user's family is obtained by comprehensively scoring the potential consumption concept of the user's family; Calculate the potential consumption level score of the user's family based on the potential consumption concept of the user's family and the broadband package information of the user's family; The potential consumption level of the user's family is obtained according to the potential consumption level score of the user's family; Identify the size of the permanent population of the user's home based on the active time period of each terminal device among the specific terminal devices used in the user's home; According to the broadband installation address of the user's family, the residential property of the user's family is obtained with the help of a real estate information platform, and the residential consumption level of the user's family is obtained based on the residential property.

6. The method according to claim 5, characterized in that The comprehensive score of the potential consumption concept of the user's family is calculated based on the specific terminal devices and quantity used in the user's family, and is specifically calculated according to the following formula: In the formula, C represents the comprehensive score of the user's family's potential consumption concept, n represents the number of specific terminal devices used in the user's family, and w i represents the weight of the i-th terminal device, C i represents the household consumption concept reflected by the i-th terminal device; The potential consumption level score of the user's family is calculated based on the potential consumption concept of the user's family and the broadband package information of the user's family, and is specifically calculated according to the following formula: L=w C ×C+w p ×P Where L represents the potential consumption level score of the user's family, w C represents the weight of the comprehensive score of potential consumption concept, P represents the score of the broadband package information of the user's family, and w p Indicates the weight of the broadband package information score.

7. The method according to claim 6, characterized in that The broadband package information score P of the user's family is calculated according to the following formula: P=α×S p +β×S s +γ×S d In the formula, S p Indicates the monthly fee of the broadband package, S s Indicates the number of additional services in the broadband package, S d represents the package period of the broadband package, α, β, and γ represent S p , S s , S d The weight it occupies.

8. The method according to claim 5, characterized in that The identifying the size of the resident population of the user's home based on the active time period of each terminal device in the specific terminal devices used in the user's home specifically includes: Taking terminal devices as units, count the active time periods of each terminal device in the user's home, and calculate the average usage time ratio of each terminal device within the preset time intervals of various APP applications; According to the active time period of each terminal device and the average usage time ratio of each terminal device within the preset time interval of each type of APP application, the usage patterns of the terminal devices are clustered to obtain the permanent population size of the user's family.

9. A device for formulating a broadband marketing strategy, characterized in that: The device comprises: A data acquisition module, used to acquire flexible packet inspection technology EDPI data corresponding to target broadband account information within a preset time interval; A quality difference score acquisition module, connected to the data acquisition module, for acquiring the network quality difference score of the user in each time period within the preset time interval according to the EDPI data; A quality difference portrait acquisition module, connected to the quality difference score acquisition module, is used to obtain the network quality change of the user in the preset time interval according to the network quality difference score of each time period, and obtain the network quality difference portrait of the user; A key dimension acquisition module, connected to the data acquisition module, for performing a user family attribute portrait analysis based on the EDPI data to obtain key dimensions for constructing an accurate user family attribute portrait, wherein the key dimensions include at least two of the user's Internet preferences, family consumption capacity and resident population, family member age distribution, and active time period; A family portrait construction module, connected to the key dimension acquisition module, for constructing an accurate portrait of the user's family attributes according to the key dimensions used to construct an accurate portrait of the user's family attributes; The marketing strategy formulation module is connected to the quality difference portrait acquisition module and the family portrait construction module, and is used to formulate a broadband marketing strategy based on the network quality difference portrait and the accurate portrait of the user's family attributes.

10. A device for formulating a broadband marketing strategy, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to implement the method for formulating a broadband marketing strategy as claimed in any one of claims 1 to 8.

11. 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 method for formulating a broadband marketing strategy according to any one of claims 1 to 8 is implemented.

Images