Content recommendation method and device based on Gaussian mixture model, equipment and medium

Abstract

The invention relates to the technical field of machine learning and financial science and technology, and particularly discloses a content recommendation method and device based on a Gaussian mixture model, computer equipment and a storage medium. When it is detected that the user logs in, user information is obtained, the user information is processed based on a first preset algorithm, and a user category is determined; based on a preset mapping relation table, an index chain recommendation template corresponding to the user category is determined, and the mapping relation table comprises a mapping relation between the user category and the index chain recommendation template; and based on the index chain recommendation template, performing content recommendation of the operating indexes for a user to view. According to the application, accurate content recommendation can be provided for the user according to the index chain recommendation template corresponding to the user category, the user does not need to spend time and energy in screening and searching interested operating indexes and billboard contents in the content of the commercial App, missing of important operating information is also avoided, and the enterprise management efficiency and decision accuracy of the user are improved.

Description

Technical Field

[0001] The present application relates to the fields of machine learning and financial technology, and in particular to a content recommendation method, device, computer equipment and storage medium based on a Gaussian mixture model. Background Art

[0002] With the increasing complexity of business operations and the improvement of the degree of informatization, business managers need to understand the business situation of the enterprise in a timely manner in order to make correct decisions and management. Business App (software application) as a common enterprise informatization tool has become an important channel for employers or executives to understand the business situation of the enterprise. Business Apps usually contain a wealth of business indicators and dashboard content to help managers better understand the sales, profits, human resources, inventory, supply chain and other aspects of the enterprise. However, with the increasing number of business indicators and dashboard content in business apps, users need to filter and find the business indicators and dashboard content they need in the content of business apps. For example, when banks are conducting business management or assessing the risk status of certain companies, business personnel need to obtain task-related business indicators and data such as operating income, profit, asset size, and asset profit rate from a large number of contents. This not only wastes time and energy and reduces the efficiency of enterprise management, but also may miss important business information, resulting in inaccurate evaluation results of related businesses, thereby reducing the accuracy of decision-making. Therefore, how to improve the efficiency of enterprise management and the accuracy of decision-making has become an urgent problem to be solved. Summary of the invention

[0003] The present application provides a content recommendation method, device, computer equipment and storage medium based on a Gaussian mixture model to improve enterprise management efficiency and decision-making accuracy.

[0004] In a first aspect, the present application provides a content recommendation method based on a Gaussian mixture model, the method comprising:

[0005] When detecting that a user has logged in, obtaining user information, and processing the user information based on a first preset algorithm to determine a user category;

[0006] Based on a preset mapping relationship table, determining an indicator chain recommendation template corresponding to the user category, wherein the mapping relationship table includes a mapping relationship between the user category and the indicator chain recommendation template;

[0007] Based on the indicator chain recommendation template, content recommendations of business indicators are made for users to view.

[0008] In a second aspect, the present application further provides a content recommendation device based on a Gaussian mixture model, the device comprising:

[0009] A user category determination module, configured to obtain user information when detecting that a user has logged in, and process the user information based on a first preset algorithm to determine the user category;

[0010] A recommendation template determination module, configured to determine an indicator chain recommendation template corresponding to the user category based on a preset mapping relationship table, wherein the mapping relationship table includes a mapping relationship between the user category and the indicator chain recommendation template;

[0011] The content recommendation module is used to recommend content of business indicators based on the indicator chain recommendation template for users to view.

[0012] In a third aspect, the present application also provides a computer device, comprising a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program and implement the content recommendation method based on the Gaussian mixture model as described above when executing the computer program.

[0013] In a fourth aspect, the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor implements the content recommendation method based on the Gaussian mixture model as described above.

[0014] The present application discloses a content recommendation method, device, computer equipment and storage medium based on a Gaussian mixture model. When a user login is detected, user information is obtained, and the user information is processed based on a first preset algorithm to determine the user category; based on a preset mapping relationship table, an indicator chain recommendation template corresponding to the user category is determined, wherein the mapping relationship table includes a mapping relationship between the user category and the indicator chain recommendation template; content recommendations of business indicators are made based on the indicator chain recommendation template for users to view. The present application can provide users with accurate content recommendations based on the indicator chain recommendation template corresponding to the user category, without the need for users to spend time and energy screening and searching for business indicators and dashboard content of interest in the content of the business app, and also avoids missing important business information, thereby improving users' business management efficiency and decision-making accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0016] Figure 1is a schematic flow chart of a first embodiment of a content recommendation method based on a Gaussian mixture model provided by an embodiment of the present application;

[0017] Figure 2 is a schematic flow chart of a second embodiment of a content recommendation method based on a Gaussian mixture model provided in an embodiment of the present application;

[0018] Figure 3 is a schematic flow chart of a third embodiment of a content recommendation method based on a Gaussian mixture model provided by an embodiment of the present application;

[0019] Figure 4 A schematic block diagram of a content recommendation device based on a Gaussian mixture model provided in an embodiment of the present application;

[0020] Figure 5 A schematic block diagram of the structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0022] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0023] It should be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0024] It should be further understood that the term “and / or” used in the specification and appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0025] The embodiments of the present application provide a content recommendation method, device, computer equipment and storage medium based on a Gaussian mixture model. Among them, the content recommendation method based on the Gaussian mixture model can be applied to a server, and according to the indicator chain recommendation template corresponding to the user category, accurate content recommendations are provided to users, without the need for users to spend time and energy screening and searching for business indicators and dashboard content of interest in the content of the business app, and also avoids missing important business information, thereby improving the user's business management efficiency and decision-making accuracy. Among them, the server can be an independent server or a server cluster.

[0026] In conjunction with the accompanying drawings, some embodiments of the present application are described in detail below. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0027] See also Figure 1 , Figure 1 This is a schematic flow chart of the first embodiment of a content recommendation method based on a Gaussian mixture model provided by an embodiment of the present application. The content recommendation method based on a Gaussian mixture model can be applied to a server to provide users with accurate content recommendations based on an indicator chain recommendation template corresponding to a user category, without requiring users to spend time and energy screening and searching for business indicators and dashboard content of interest in the content of a business app, thereby avoiding the omission of important business information and improving the user's business management efficiency and decision-making accuracy.

[0028] like Figure 1 As shown, the content recommendation method based on the Gaussian mixture model specifically includes steps S101 to S103.

[0029] S101, when detecting that a user has logged in, obtaining user information, and processing the user information based on a first preset algorithm to determine a user category;

[0030] In one embodiment, user information can be obtained from the user's logged-in account, or collected through pop-up window inquiries, etc. Exemplarily, user information may include user identity (such as bank credit manager, bank risk control specialist, enterprise sales manager, enterprise production manager, etc.), user-in-charge business (such as loan approval, loan risk assessment, sales strategy formulation, production strategy formulation, etc.).

[0031] In one embodiment, the first preset algorithm may be a Gaussian mixture model or other crowd clustering algorithms, which is used to group users or crowds according to their characteristics, behaviors or other relevant attributes. For example, users may be classified by user identity and the business they are in charge of, and user categories include loan risk assessors, sales strategy developers, etc.

[0032] In a specific embodiment, after the user logs in, cluster analysis is performed on the user information based on a first preset algorithm to determine the user category of the current user.

[0033] S102: determining an indicator chain recommendation template corresponding to the user category based on a preset mapping relationship table, wherein the mapping relationship table includes a mapping relationship between the user category and the indicator chain recommendation template;

[0034] In one embodiment, the mapping relationship table is used to store user categories, indicator chain recommendation templates, and the mapping relationship between each user category and the indicator chain recommendation template.

[0035] In one embodiment, the indicator chain refers to connecting multiple indicators in series in a certain order to form a chain that can reflect the trend of business changes. The indicator chain recommendation template is a standardized framework for building and analyzing indicator chains, which provides a series of indicator combinations and sequences based on business scenarios and goals.

[0036] In a specific embodiment, the mapping relationship table is searched to see whether the user category described by the current user exists. If so, the indicator chain recommendation template corresponding to the user category is obtained, so as to recommend content to the user based on the indicator chain recommendation template in the subsequent process. If not, the user's current observation data is obtained, and the current observation data is processed using a Gaussian mixture model to obtain the indicator chain recommendation template corresponding to the new category to which the user belongs, and the indicator chain recommendation template is used to recommend content to the user.

[0037] S103: Recommending content of business indicators based on the indicator chain recommendation template for users to view.

[0038] In one embodiment, according to the indicators and the indicator sequence in the indicator chain recommendation template, the corresponding indicators and the data corresponding to the indicators are obtained from the APP logged in by the user and recommended to the user for viewing.

[0039] Exemplarily, the indicators and indicator sequence in the indicator chain recommendation template may be "lender's credit status - lender's income status - lender's debt status - lender's asset status", and content is recommended to the bank credit manager according to the indicators and indicator sequence to quickly provide the bank credit manager with effective information.

[0040] The above-mentioned embodiment provides a content recommendation method, apparatus, computer equipment and storage medium based on a Gaussian mixture model, which obtains user information when a user login is detected, and processes the user information based on a first preset algorithm to determine the user category; based on a preset mapping relationship table, determines the indicator chain recommendation template corresponding to the user category, wherein the mapping relationship table includes a mapping relationship between the user category and the indicator chain recommendation template; and recommends content of business indicators based on the indicator chain recommendation template for users to view. The present application can provide users with accurate content recommendations based on the indicator chain recommendation template corresponding to the user category, without the user having to spend time and energy screening and searching for business indicators and dashboard content of interest in the content of the business app, and also avoids missing important business information, thereby improving the user's business management efficiency and decision-making accuracy.

[0041] See also Figure 2 , Figure 2 This is a schematic flow chart of the second embodiment of a content recommendation method based on a Gaussian mixture model provided in an embodiment of the present application. The content recommendation method based on a Gaussian mixture model can be applied to a server to generate a mapping relationship table between user categories and indicator chain recommendation templates based on user information and historical observation indicators of historical users, determine the indicator chain recommendation template corresponding to the user based on the mapping relationship table, and then provide accurate content recommendations for users based on the indicator chain recommendation template, without the need for users to spend time and energy screening and searching for business indicators and dashboard content of interest in the content of the business app, and also avoids missing important business information, thereby improving the user's enterprise management efficiency and decision-making accuracy.

[0042] like Figure 2 As shown, before step S103, the content recommendation method based on the Gaussian mixture model further includes steps S201 to S204.

[0043] S201, obtaining user information and historical observation indicators of at least one historical user;

[0044] In one embodiment, historical users are users who have used the current business APP, and historical observation indicators are obtained by analyzing the user's behavior data during the use of the business APP. Specifically, the user's behavior data is analyzed to extract the business indicators (such as user activity, order volume, sales, etc.) that the user has viewed as historical observation indicators.

[0045] S202, converting the historical observation index into a feature vector;

[0046] Furthermore, the step S202 includes: screening the historical observation indicators for abnormal indicators to obtain an observation indicator set; processing the observation indicator set based on a preset dimension to obtain a target indicator chain; and performing word vector conversion on the target indicator chain to obtain the feature vector.

[0047] Furthermore, the processing of the observed indicator set based on the preset dimension to obtain the target indicator chain includes: analyzing the observed indicator set based on the preset dimension to generate an initial indicator chain; and selecting each operating indicator in the initial indicator chain based on preset rules to obtain the target indicator chain.

[0048] In one embodiment, anomaly detection is performed on historical observation indicators, abnormal indicators are determined, and abnormal indicators are deleted to obtain an observation indicator set.

[0049] In one embodiment, the preset dimensions include a session dimension and a day dimension. Among them, the session dimension focuses on the user's behavior pattern in a single interaction, including the user's behavior sequence, dwell time, event triggering frequency, etc. in a session; analyzes the user's conversion path in the session, identifies key behavior nodes, such as page browsing, adding items to the shopping cart, completing purchases, etc.; identifies abnormal behaviors in the session, such as sudden jumps or long periods of inactivity. The day dimension focuses on the user's behavior pattern within a certain time period (such as a day), including the number of daily active users, user retention rate, average daily transaction amount, etc.; analyzes the daily trend and periodic changes of user behavior, such as the difference between weekdays and weekends, traffic peaks on holidays, etc.; identifies user behavior patterns on specific dates, and evaluates the impact of marketing activities or special events.

[0050] In one embodiment, the set of observed indicators is processed through the session dimension and the day dimension to obtain an indicator chain. Specifically, the data is grouped according to the session dimension, and the indicator values ​​in each session are counted, such as the number of active users, order volume, sales volume, etc. in each session. The data is grouped according to the day dimension, and the data indicator values ​​of each day are counted, such as the number of active users, order volume, sales volume, etc. every day. According to business needs, appropriate indicators are selected and connected in series in a certain order to form an indicator chain. For example, the three indicators of the number of active users, order volume, and sales volume can be selected and connected in series in chronological order to form an initial indicator chain that reflects the trend of business changes.

[0051] In one embodiment, the following formula is used to remove the repeated indexes i at adjacent positions in the index chain.

[0052]

[0053] In one embodiment, abnormal indicators in the initial indicator chain are removed, and indicators with poor timeliness (historical indicator names) are mapped to the latest operating indicator names to obtain the target indicator chain.

[0054] In one embodiment, the target indicator chain is converted into a word vector to obtain a feature vector. Specifically, by building an Embedding, each target indicator chain is converted into a feature vector that conforms to the semantic relationship of the words, and the input required by the model is obtained: Input = Embedding ([ind1, ind2, ... indn]).

[0055] S203: Process the user information and the feature vector based on a Gaussian mixture model to determine at least one user category and an indicator chain recommendation template corresponding to each user category;

[0056] Furthermore, before step S203, it also includes: based on a second preset algorithm and the historical observation indicators, obtaining target parameter data related to the Gaussian mixture model; based on the target parameter data, generating an initial Gaussian model; obtaining a verification data set, and based on the verification data set, obtaining the classification accuracy of the initial Gaussian model; when the classification accuracy is greater than or equal to a preset accuracy threshold, using the initial Gaussian model as the Gaussian mixture model.

[0057] Furthermore, the verification data set includes indicator observation data and an actual observation indicator chain of at least one user, and obtaining the classification accuracy of the initial Gaussian model based on the verification data set includes: processing the indicator observation data of at least one user based on the initial Gaussian model to obtain a prediction indicator chain corresponding to at least one of the users; comparing the prediction indicator chain with the actual indicator chain to obtain the classification accuracy.

[0058] In one embodiment, the second preset algorithm is the EM (Expectation-Maximization Algorithm) algorithm. Specifically, assuming that there are M Gaussian distributions in the indicator chain latent space, the probability distribution obeyed by the indicator chain is:

[0059]

[0060] Where P(x) is the probability that the observed data x belongs to the current classification, p(x|m)=p(x; μ m ,Σm) is the probability density function of the mth Gaussian model, p(k)=π m is the weight of the mth Gaussian model, Σm is the covariance of the mth Gaussian model, μ m is the mean of the mth Gaussian model.

[0061] The EM algorithm is used to iteratively solve the problem, which is divided into E step and M step, as follows:

[0062] Step E:

[0063]

[0064] Among them, z i =j means that the i-th indicator comes from the j-th model, is the probability that the i-th indicator comes from the j-th model.

[0065] The M step is:

[0066]

[0067] in, The i-th index is from z (i) The probability of z (i) is the category corresponding to the i-th indicator.

[0068] By solving the Laplace method, we can get:

[0069]

[0070] make Then there are:

[0071]

[0072] in,

[0073] When the iteration is completed, the target parameter data π, μ, Σ related to the Gaussian mixture model are obtained.

[0074] In one embodiment, an initial Gaussian model is obtained according to the target parameter data, and the indicator observation data in the verification data set are clustered accordingly by the initial Gaussian model to obtain a prediction indicator chain for each type of user. The prediction indicator chain is compared with the actual indicator chain in the verification data set to obtain the classification accuracy of the initial Gaussian model.

[0075] In one embodiment, when the classification accuracy is greater than or equal to the preset accuracy threshold, the initial Gaussian model at this time is used as the Gaussian mixture model. When the classification accuracy is less than the preset accuracy threshold, the relevant parameters of the model are adjusted, such as increasing or decreasing the number of Gaussian distributions (i.e., the number of components), to obtain a new Gaussian model until the classification accuracy is greater than or equal to the preset accuracy threshold. The accuracy threshold can be freely set by the user as needed.

[0076] S204: Generate the mapping relationship table based on at least one user category and the indicator chain recommendation template corresponding to each user category.

[0077] In one embodiment, the correspondence between each user category and the indicator chain recommendation template is used as the mapping relationship between the user category and the indicator chain recommendation template.

[0078] In one embodiment, each user category and its corresponding indicator chain recommendation template are stored in a table in a one-to-one correspondence to generate a mapping relationship table.

[0079] See also Figure 3 , Figure 3 This is a schematic flow chart of the third embodiment of a content recommendation method based on a Gaussian mixture model provided by an embodiment of the present application. The content recommendation method based on a Gaussian mixture model can be applied to a server, and is used to obtain the user's current observation data when the user is a new category user, obtain the indicator chain recommendation template of the new category user through the Gaussian mixture model, and recommend content of business indicators to the user, which can improve the observation efficiency of the new category user in the business APP, thereby improving the user's enterprise management efficiency and decision-making accuracy.

[0080] like Figure 3 As shown, after step S101, the content recommendation method based on the Gaussian mixture model further includes steps S301 to S303.

[0081] S301, when the user category is a new category, monitoring current observation data of the user;

[0082] S302: Analyze the current observation data based on a Gaussian mixture model to obtain an indicator chain recommendation template corresponding to the new category;

[0083] S303: Recommending content of business indicators to the user based on the indicator chain recommendation template.

[0084] In one embodiment, when the user category is a new category, it is impossible to determine the indicator chain recommendation template corresponding to the new category according to the preset mapping relationship table. It is necessary to monitor the user's operation and obtain the user's current observation data.

[0085] In one embodiment, the currently observed data is the indicator data that the new category user is currently observing.

[0086] In one embodiment, the current observation data is analyzed using a Gaussian mixture model to obtain an indicator chain recommendation template corresponding to the new category. Specifically, anomaly detection is performed on the current observation data and the abnormal data is deleted. The current observation data after the abnormal data is deleted is processed based on the session dimension and the day dimension to generate an initial indicator chain. The duplicate indicators and abnormal indicators in adjacent positions in the initial indicator chain are eliminated to obtain a target indicator chain. The target indicator chain is converted into a feature vector that conforms to the semantic relationship of the words through Embedding as an input parameter of the Gaussian mixture model. The feature vector is input into the Gaussian mixture model to obtain an indicator recommendation template for the new category of users.

[0087] In one embodiment, content recommendations can be made during the user's current observation process based on the obtained indicator recommendation template. Exemplarily, according to the user's current observation process, the user's current observation data is obtained, and after the indicator chain recommendation template is obtained according to the current observation data, subsequent observation content is recommended to the user. This can improve the observation efficiency of new categories of users in business apps, thereby improving the user's enterprise management efficiency and decision-making accuracy.

[0088] In another embodiment, content recommendation may also be performed during the next observation of the user based on the obtained indicator recommendation template.

[0089] In one embodiment, after obtaining the indicator chain recommendation template for a new category of users, the user category and the indicator recommendation template are stored in a mapping relationship table for next use.

[0090] In the above embodiment, when the user is a new category user, the user's current observation data can be obtained, and the indicator chain recommendation template of the new category user can be obtained through the Gaussian mixture model, and the content of business indicators can be recommended to the user, which can improve the observation efficiency of the new category users in the business APP, thereby improving the user's enterprise management efficiency and decision-making accuracy.

[0091] See also Figure 4 , Figure 4 The embodiment of the present application provides a schematic block diagram of a content recommendation device based on a Gaussian mixture model, and the content recommendation device based on a Gaussian mixture model is used to execute the aforementioned content recommendation method based on a Gaussian mixture model. The content recommendation device based on a Gaussian mixture model can be configured on a server.

[0092] like Figure 4 As shown, the content recommendation device 400 based on the Gaussian mixture model includes:

[0093] The user category determination module 401 is used to obtain user information when detecting that a user is logged in, and process the user information based on a first preset algorithm to determine the user category;

[0094] A recommended template determination module 402 is used to determine the indicator chain recommendation template corresponding to the user category based on a preset mapping relationship table, wherein the mapping relationship table includes a mapping relationship between the user category and the indicator chain recommendation template;

[0095] The content recommendation module 403 is used to recommend content of business indicators based on the indicator chain recommendation template for users to view.

[0096] Furthermore, the Gaussian mixture model-based content recommendation device 400 further includes a mapping relationship table generation module, and the mapping relationship table generation module includes:

[0097] A historical data acquisition submodule, used to acquire user information and historical observation indicators of at least one historical user;

[0098] A feature vector acquisition submodule, used to convert the historical observation indicators into feature vectors;

[0099] A recommendation template acquisition submodule, used to process the user information and the feature vector based on a Gaussian mixture model to determine at least one user category and an indicator chain recommendation template corresponding to each user category;

[0100] The mapping relationship table generation submodule is used to generate the mapping relationship table based on at least one user category and the indicator chain recommendation template corresponding to each user category.

[0101] Furthermore, the feature vector acquisition submodule includes:

[0102] An indicator set obtaining unit, used to screen the historical observation indicators for abnormal indicators and obtain an observation indicator set;

[0103] A target indicator chain obtaining unit, used to process the observed indicator set based on a preset dimension to obtain a target indicator chain;

[0104] The feature vector obtaining unit is used to perform word vector conversion on the target indicator chain to obtain the feature vector.

[0105] Furthermore, the target indicator chain obtaining unit includes:

[0106] An initial indicator chain generating subunit, used for analyzing the observed indicator set based on a preset dimension to generate an initial indicator chain;

[0107] The target indicator chain obtaining subunit is used to select each operating indicator in the initial indicator chain based on preset rules to obtain the target indicator chain.

[0108] Furthermore, the recommended template determination module 402 further includes a Gaussian mixture model acquisition submodule, and the Gaussian mixture model acquisition submodule includes:

[0109] A target parameter data obtaining unit, configured to obtain target parameter data related to the Gaussian mixture model based on a second preset algorithm and the historical observation index;

[0110] An initial Gaussian model generating unit, used to generate an initial Gaussian model based on the target parameter data;

[0111] A classification accuracy obtaining unit, used to obtain a verification data set, and obtain the classification accuracy of the initial Gaussian model based on the verification data set;

[0112] A Gaussian mixture model determination unit is used to use the initial Gaussian model as the Gaussian mixture model when the classification accuracy is greater than or equal to a preset accuracy threshold.

[0113] Furthermore, the verification data set includes at least one user's indicator observation data and an actual observation indicator chain, and the classification accuracy acquisition unit includes:

[0114] An observation indicator chain obtaining subunit, configured to process the indicator observation data of at least one user based on the initial Gaussian model to obtain a prediction indicator chain corresponding to at least one of the users;

[0115] The classification accuracy obtaining subunit is used to compare the predicted indicator chain with the actual indicator chain to obtain the classification accuracy.

[0116] Furthermore, the Gaussian mixture model-based content recommendation device 400 further includes:

[0117] A data monitoring module, used for monitoring the current observation data of the user when the user category is a new category;

[0118] A template acquisition module, used to analyze the current observation data based on a Gaussian mixture model to obtain an indicator chain recommendation template corresponding to the new category;

[0119] A content recommendation module is used to recommend content of business indicators to the user based on the indicator chain recommendation template.

[0120] It should be noted that those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described device and each module can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0121] The above-mentioned device can be implemented in the form of a computer program. Figure 5 Runs on the computer device shown.

[0122] See also Figure 5 , Figure 5 1 is a schematic block diagram of a computer device provided in an embodiment of the present application. The computer device may be a server.

[0123] See also Figure 5 The computer device includes a processor, a memory and a network interface connected through a system bus, wherein the memory may include a non-volatile storage medium and an internal memory.

[0124] The non-volatile storage medium can store an operating system and a computer program. The computer program includes program instructions, and when the program instructions are executed, the processor can execute any content recommendation method based on a Gaussian mixture model.

[0125] The processor is used to provide computing and control capabilities and support the operation of the entire computer equipment.

[0126] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor, the processor can execute any content recommendation method based on the Gaussian mixture model.

[0127] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art will appreciate that Figure 5 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0128] It should be understood that the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0129] In one embodiment, the processor is used to run a computer program stored in the memory to implement the following steps:

[0130] When detecting that a user has logged in, obtaining user information, and processing the user information based on a first preset algorithm to determine a user category;

[0131] Based on a preset mapping relationship table, determining an indicator chain recommendation template corresponding to the user category, wherein the mapping relationship table includes a mapping relationship between the user category and the indicator chain recommendation template;

[0132] Based on the indicator chain recommendation template, content recommendations of business indicators are made for users to view.

[0133] In one embodiment, before determining the indicator chain recommendation template corresponding to the user category based on the preset mapping relationship table, the processor is further configured to implement:

[0134] Obtain user information and historical observation indicators of at least one historical user;

[0135] Convert the historical observation indicators into feature vectors;

[0136] Based on a Gaussian mixture model, the user information and the feature vector are processed to determine at least one user category and an indicator chain recommendation template corresponding to each user category;

[0137] The mapping relationship table is generated based on at least one user category and an indicator chain recommendation template corresponding to each user category.

[0138] In one embodiment, when converting the historical observation indicator into a feature vector, the processor is used to implement:

[0139] Screening the historical observation indicators for abnormal indicators to obtain an observation indicator set;

[0140] Processing the observed indicator set based on a preset dimension to obtain a target indicator chain;

[0141] The target indicator chain is subjected to word vector conversion to obtain the feature vector.

[0142] In one embodiment, when the processor processes the observed indicator set based on the preset dimension to obtain the target indicator chain, it is used to implement:

[0143] Analyze the set of observed indicators based on preset dimensions to generate an initial indicator chain;

[0144] Based on preset rules, each operating indicator in the initial indicator chain is selected to obtain a target indicator chain.

[0145] In one embodiment, before the processor processes the user information and the feature vector based on a Gaussian mixture model to determine at least one user category and an indicator chain recommendation template corresponding to each user category, it is further configured to implement:

[0146] Based on a second preset algorithm and the historical observation indicators, obtaining target parameter data related to the Gaussian mixture model;

[0147] Based on the target parameter data, generating an initial Gaussian model;

[0148] Acquire a validation data set, and based on the validation data set, obtain the classification accuracy of the initial Gaussian model;

[0149] When the classification accuracy is greater than or equal to a preset accuracy threshold, the initial Gaussian model is used as the Gaussian mixture model.

[0150] In one embodiment, the verification data set includes indicator observation data of at least one user and an actual observation indicator chain. When the processor obtains the classification accuracy of the initial Gaussian model based on the verification data set, it is used to implement:

[0151] Based on the initial Gaussian model, the indicator observation data of at least one user is processed to obtain a prediction indicator chain corresponding to at least one of the users;

[0152] The predicted indicator chain is compared with the actual indicator chain to obtain the classification accuracy.

[0153] In one embodiment, the processor is further configured to implement, when detecting that a user is logged in, obtaining user information, and processing the user information based on a first preset algorithm to determine the user category:

[0154] When the user category is a new category, monitoring current observation data of the user;

[0155] Based on the Gaussian mixture model, the current observation data is analyzed to obtain an indicator chain recommendation template corresponding to the new category;

[0156] Based on the indicator chain recommendation template, content of business indicators is recommended to the user.

[0157] A computer-readable storage medium is also provided in an embodiment of the present application, wherein the computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and the processor executes the program instructions to implement any one of the content recommendation methods based on the Gaussian mixture model provided in the embodiments of the present application.

[0158] The computer-readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiment, such as a hard disk or memory of the computer device. The computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, a smart memory card (SmartMedia Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc., equipped on the computer device.

[0159] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present application, and these modifications or replacements should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A content recommendation method based on a Gaussian mixture model, characterized in that: include: When detecting a user login, obtaining user information, and processing the user information based on a first preset algorithm to determine the user category; Determining an indicator chain recommendation template corresponding to the user category based on a preset mapping relationship table, wherein the mapping relationship table includes a mapping relationship between the user category and the indicator chain recommendation template; Based on the indicator chain recommendation template, content recommendations of business indicators are made for users to view.

2. The content recommendation method based on Gaussian mixture model according to claim 1, characterized in that: Before determining the indicator chain recommendation template corresponding to the user category based on the preset mapping relationship table, the method further includes: Obtain user information and historical observation indicators of at least one historical user; Converting the historical observation indicators into feature vectors; Processing the user information and the feature vector based on a Gaussian mixture model to determine at least one user category and an indicator chain recommendation template corresponding to each user category; The mapping relationship table is generated based on at least one user category and an indicator chain recommendation template corresponding to each user category.

3. The content recommendation method based on Gaussian mixture model according to claim 2, characterized in that: The converting the historical observation indicator into a feature vector includes: Screening the historical observation indicators for abnormal indicators to obtain an observation indicator set; Processing the observed indicator set based on preset dimensions to obtain a target indicator chain; Perform word vector conversion on the target indicator chain to obtain the feature vector.

4. The content recommendation method based on Gaussian mixture model according to claim 3, characterized in that: The step of processing the set of observed indicators based on the preset dimensions to obtain a target indicator chain includes: Analyze the set of observed indicators based on preset dimensions to generate an initial indicator chain; Based on preset rules, each operating indicator in the initial indicator chain is selected to obtain a target indicator chain.

5. The content recommendation method based on Gaussian mixture model according to claim 2, characterized in that: Before processing the user information and the feature vector based on the Gaussian mixture model to determine at least one user category and an indicator chain recommendation template corresponding to each user category, the method further includes: Based on a second preset algorithm and the historical observation indicators, obtaining target parameter data related to the Gaussian mixture model; Based on the target parameter data, generating an initial Gaussian model; Obtaining a validation data set, and obtaining a classification accuracy of the initial Gaussian model based on the validation data set; When the classification accuracy is greater than or equal to a preset accuracy threshold, the initial Gaussian model is used as the Gaussian mixture model.

6. The content recommendation method based on Gaussian mixture model according to any one of claim 5, characterized in that The validation data set includes indicator observation data of at least one user and an actual observation indicator chain. The obtaining of the classification accuracy of the initial Gaussian model based on the validation data set includes: Based on the initial Gaussian model, processing indicator observation data of at least one user to obtain a prediction indicator chain corresponding to at least one of the users; The predicted indicator chain is compared with the actual indicator chain to obtain the classification accuracy.

7. The content recommendation method based on Gaussian mixture model according to claims 1 to 6, characterized in that: When detecting that a user has logged in, obtaining user information, and processing the user information based on a first preset algorithm to determine the user category, further comprising: When the user category is a new category, monitoring current observation data of the user; Analyze the current observation data based on a Gaussian mixture model to obtain an indicator chain recommendation template corresponding to the new category; Based on the indicator chain recommendation template, content of business indicators is recommended to the user.

8. A content recommendation device based on a Gaussian mixture model, characterized in that: include: A user category determination module is used to obtain user information when detecting a user login, and process the user information based on a first preset algorithm to determine the user category; A recommendation template determination module, configured to determine an indicator chain recommendation template corresponding to the user category based on a preset mapping relationship table, wherein the mapping relationship table includes a mapping relationship between the user category and the indicator chain recommendation template; The content recommendation module is used to recommend content of business indicators based on the indicator chain recommendation template for users to view.

9. A computer device, characterized in that: The computer device includes a memory and a processor; The memory is used to store computer programs; The processor is configured to execute the computer program and implement the content recommendation method based on the Gaussian mixture model according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which, when executed by a processor, enables the processor to implement the content recommendation method based on a Gaussian mixture model according to any one of claims 1 to 7.

Images