A multi-terminal meal ordering intelligent management method and system based on multiple users
By using the DBSCAN algorithm and decision tree model in a multi-terminal food ordering system, the problems of user segmentation and data applicability analysis are solved, and accurate and personalized catering recommendations are achieved, which improves user experience and recommendation efficiency.
Patent Information
- Application Number
- CN202411011441.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-26
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-07-26
AI Technical Summary
The existing multi-terminal ordering system lacks segmentation and personalized recommendations for user types, resulting in poor user experience and lack of applicability analysis for multi-terminal data, affecting the user experience of catering recommendations.
The target catering platform collects user ordering interaction data, uses the DBSCAN algorithm to cluster and group users, combines the catering dish database to perform recommendation analysis, and builds a classification model based on the decision tree to generate multi-terminal data recommendation solutions.
It realizes accurate and personalized catering data recommendation, improves user satisfaction and recommendation efficiency, and is suitable for dynamic data recommendation under multiple terminals.
Smart Images

Figure CN118536951B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of intelligent catering data analysis, and more specifically, to an intelligent management method and system for multi-terminal meal ordering based on multiple users. Background Art
[0002] With the continuous development of information technology, the way of ordering food has gradually changed from traditional paper menu ordering to multi-terminal intelligent ordering. However, the existing multi-terminal ordering system often lacks segmentation and personalized recommendations for user types, resulting in poor user experience. At the same time, the existing technology lacks data applicability analysis for multi-terminal display, which affects the user experience based on catering recommendations. Therefore, there is an urgent need for a multi-terminal ordering intelligent management method based on multiple users. Summary of the invention
[0003] The present invention overcomes the defects of the prior art and proposes a multi-terminal meal ordering intelligent management method and system based on multiple users.
[0004] The first aspect of the present invention provides a method for intelligent management of multi-terminal meal ordering based on multiple users, comprising:
[0005] Through the target catering platform, collect the user ordering interaction data in the historical time period, and perform user feature analysis and feature extraction on multiple users based on the interaction data to form user feature data;
[0006] A clustering model is built based on the DBSCAN algorithm, and user feature data is imported into the clustering model as clustering sample data for clustering grouping. Based on the clustering results, the user group is divided into multiple user groups;
[0007] Based on the target catering platform, the current user interaction information is obtained, and user feature analysis and extraction are performed based on the current user interaction information to obtain the current user feature data, and the clustering group of the current user is screened in combination with the clustering model to obtain the user group to which the current user belongs;
[0008] Collect user feature data within the user group to form a current feature data set, and perform catering data recommendation analysis based on the current feature data set in the catering menu database to form a recommendation data set;
[0009] Obtain display demand information of N preset terminals, perform data format condition analysis based on the display demand information, and form N condition feature data, build a classification model based on a decision tree, use the N condition feature data as judgment nodes of the classification model, and build a classification model that forms a complete decision tree;
[0010] The recommendation data set is imported into the classification model for data classification. Based on the classification results, the recommendation data set is divided into multiple groups of recommendation data. Based on the display demand information of N preset terminals, the multiple groups of recommendation data are respectively associated with the preset terminals, and a multi-terminal data recommendation plan is generated based on the associated information and the multiple groups of recommendation data.
[0011] In this solution, the target catering platform collects user ordering interaction data in the historical time period, and performs user feature analysis and feature extraction on multiple users based on the interaction data to form user feature data, specifically:
[0012] Based on a preset historical time period, the user ordering interaction data of all users is collected through the target catering platform;
[0013] Through the user ordering interaction data, user feature analysis and feature extraction are performed on various users to form user feature data of various users;
[0014] The user characteristic analysis includes dimensions of dish type, ingredient variety, order time, ordering frequency, and consumption amount information.
[0015] In this solution, a clustering model is constructed based on the DBSCAN algorithm, user feature data is imported into the clustering model as clustering sample data for clustering grouping, and the user group is divided into multiple user groups based on the clustering results, specifically:
[0016] Construct a clustering model based on the DBSCAN algorithm;
[0017] The user feature data is associated with the user information, and the user feature data is imported into the clustering model as the clustering sample data. The minimum distance threshold and the minimum number of neighbors are set for data classification and clustering expansion, and finally the clustering results are obtained based on the data set division.
[0018] Based on the clustering results, the corresponding user groups are divided into related groups to obtain multiple user groups.
[0019] In this solution, the current user interaction information is obtained based on the target catering platform, user feature analysis and extraction are performed based on the current user interaction information to obtain the current user feature data, and the clustering group of the current user is screened in combination with the clustering model, and the user group to which the current user belongs is obtained, specifically:
[0020] Analyze real-time user information through the target catering platform, obtain real-time user interaction information and mark it as current user interaction information;
[0021] Perform user feature analysis and feature extraction based on current user interaction information to form current user feature data;
[0022] The current user feature data is imported into the clustering model and the distance to each clustering group is calculated, and the user group to which the current user feature data belongs is determined by the shortest distance.
[0023] In this solution, the user feature data in the user group is collected to form a current feature data set, and in the catering menu database, catering data recommendation analysis is performed based on the current feature data set to form a recommendation data set, specifically:
[0024] Through the clustering model, the user feature data within the user group is collected and integrated to form the current feature data set;
[0025] A collaborative filtering recommendation algorithm is introduced, and the current feature data set is used as the user interest feature. The recommendation data is analyzed and screened in the restaurant menu database to obtain the recommendation data set.
[0026] In this solution, the display demand information of N preset terminals is obtained, data format condition analysis is performed through the display demand information, and N condition feature data are formed, a classification model based on a decision tree is constructed, and the N condition feature data are used as judgment nodes of the classification model to construct a classification model forming a complete decision tree, specifically:
[0027] Obtain display demand information of N preset terminals, analyze the adapted data format and data type based on the display demand information, and generate corresponding data condition information;
[0028] Extract feature information based on data condition information to form N types of conditional feature data;
[0029] Build a decision tree-based classification model;
[0030] Based on N kinds of conditional feature data, feature condition conversion based on decision tree nodes is performed to form N conditional nodes;
[0031] Taking N conditional nodes as the initial nodes in the classification model, the classification model is determined based on the heuristic algorithm to determine each node and classification condition, and a classification model based on the complete decision tree is constructed.
[0032] In this solution, the recommendation data set is imported into the classification model for data classification. Based on the classification result, the recommendation data set is divided into multiple groups of recommendation data. Based on the display demand information of N preset terminals, the multiple groups of recommendation data are respectively associated with the preset terminals. Based on the association information and the multiple groups of recommendation data, a multi-terminal data recommendation solution is generated, which is specifically as follows:
[0033] Import the recommended data set into the classification model for data classification and obtain the classification results;
[0034] Based on the classification results, the recommendation data set is divided into multiple groups of recommendation data;
[0035] Based on display demand information of N preset terminals, multiple groups of recommended data are associated with the preset terminals respectively, and a multi-terminal data recommendation scheme is generated based on the associated information and the multiple groups of recommended data.
[0036] The second aspect of the present invention further provides a multi-terminal meal ordering intelligent management system based on multiple users, the system comprising: a memory, a processor, the memory comprising a multi-terminal meal ordering intelligent management program based on multiple users, the multi-terminal meal ordering intelligent management program based on multiple users when executed by the processor implements the following steps:
[0037] Through the target catering platform, collect the user ordering interaction data in the historical time period, and perform user feature analysis and feature extraction on multiple users based on the interaction data to form user feature data;
[0038] A clustering model is built based on the DBSCAN algorithm, and user feature data is imported into the clustering model as clustering sample data for clustering grouping. Based on the clustering results, the user group is divided into multiple user groups;
[0039] Based on the target catering platform, the current user interaction information is obtained, and user feature analysis and extraction are performed based on the current user interaction information to obtain the current user feature data, and the clustering group of the current user is screened in combination with the clustering model to obtain the user group to which the current user belongs;
[0040] Collect user feature data within the user group to form a current feature data set, and perform catering data recommendation analysis based on the current feature data set in the catering menu database to form a recommendation data set;
[0041] Obtain display demand information of N preset terminals, perform data format condition analysis based on the display demand information, and form N condition feature data, build a classification model based on a decision tree, use the N condition feature data as judgment nodes of the classification model, and build a classification model that forms a complete decision tree;
[0042] The recommendation data set is imported into the classification model for data classification. Based on the classification results, the recommendation data set is divided into multiple groups of recommendation data. Based on the display demand information of N preset terminals, the multiple groups of recommendation data are respectively associated with the preset terminals, and a multi-terminal data recommendation plan is generated based on the associated information and the multiple groups of recommendation data.
[0043] The third aspect of the present invention also provides a computer-readable storage medium, which includes an intelligent management program for multi-terminal ordering based on multiple users. When the intelligent management program for multi-terminal ordering based on multiple users is executed by a processor, the steps of the intelligent management method for multi-terminal ordering based on multiple users as described in any one of the above items are implemented.
[0044] The present invention discloses an intelligent management method and system for multi-terminal ordering based on multiple users. The present invention relates to the field of catering recommendation technology. The user ordering interaction data is collected through the target catering platform, and the users are clustered and grouped using the DBSCAN algorithm. Based on the current user interaction information, the user group to which they belong is matched, and the user feature data in the group is collected, and recommendation analysis is performed in combination with the catering menu database. According to the display requirements of N preset terminals, a classification model based on a decision tree is constructed, the recommended data set is classified, and a multi-terminal data recommendation scheme is generated. The present invention realizes accurate and personalized catering data recommendation, improves user satisfaction and recommendation efficiency, and is suitable for dynamic data recommendation under multiple terminals. BRIEF DESCRIPTION OF THE DRAWINGS
[0045] Figure 1 A flowchart of a multi-terminal meal ordering intelligent management method based on multiple users of the present invention is shown;
[0046] Figure 2 A block diagram of a multi-terminal meal ordering intelligent management system based on multiple users of the present invention is shown. DETAILED DESCRIPTION
[0047] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.
[0048] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.
[0049] Figure 1 The flowchart of the present invention shows a method for intelligent management of multi-terminal meal ordering based on multiple users.
[0050] like Figure 1 As shown, the first aspect of the present invention provides a multi-terminal meal ordering intelligent management method based on multiple users, comprising:
[0051] S102, collecting user ordering interaction data within a historical time period through the target catering platform, and performing user feature analysis and feature extraction on multiple users based on the interaction data to form user feature data;
[0052] S104, building a clustering model based on the DBSCAN algorithm, importing the user feature data as clustering sample data into the clustering model for clustering grouping, and dividing the user group into multiple user groups based on the clustering results;
[0053] S106, obtaining the current user interaction information based on the target catering platform, performing user feature analysis and extraction based on the current user interaction information, obtaining the current user feature data, screening the clustering group of the current user in combination with the clustering model, and obtaining the user group to which the current user belongs;
[0054] S108, collecting user feature data in the user group to form a current feature data set, and performing catering data recommendation analysis based on the current feature data set in the catering menu database to form a recommendation data set;
[0055] S110, obtaining display demand information of N preset terminals, performing data format condition analysis based on the display demand information, and forming N types of conditional feature data, constructing a classification model based on a decision tree, using the N types of conditional feature data as judgment nodes of the classification model, and constructing a classification model that forms a complete decision tree;
[0056] S112, importing the recommendation data set into the classification model for data classification, based on the classification result, dividing the recommendation data set into multiple groups of recommendation data, based on the display demand information of N preset terminals, the multiple groups of recommendation data are respectively associated with the preset terminals, and a multi-terminal data recommendation scheme is generated based on the associated information and the multiple groups of recommendation data.
[0057] According to an embodiment of the present invention, the user ordering interaction data in the historical time period is collected through the target catering platform, and user feature analysis and feature extraction are performed on multiple users based on the interaction data to form user feature data, specifically:
[0058] Based on a preset historical time period, the user ordering interaction data of all users is collected through the target catering platform;
[0059] Through the user ordering interaction data, user feature analysis and feature extraction are performed on various users to form user feature data of various users;
[0060] The user characteristic analysis includes dimensions of dish type, ingredient variety, order time, ordering frequency, and consumption amount information.
[0061] It should be noted that the target catering platform includes a user data module, an order module, a database module, a catering interaction module, a display module, etc., which are responsible for different catering information service functions. The historical time period can be the historical operation time period of the platform, or it can be a time period set by the user, specifically a longer time period, which is used to collect enough user data for feature analysis. The user interaction data is the data collected when the user orders food or browses data on various terminals.
[0062] According to an embodiment of the present invention, the clustering model is constructed based on the DBSCAN algorithm, user feature data is imported into the clustering model as clustering sample data for clustering grouping, and the user group is divided into multiple user groups based on the clustering results, specifically:
[0063] Construct a clustering model based on the DBSCAN algorithm;
[0064] The user feature data is associated with the user information, and the user feature data is imported into the clustering model as the clustering sample data. The minimum distance threshold and the minimum number of neighbors are set for data classification and clustering expansion, and finally the clustering results are obtained based on the data set division.
[0065] Based on the clustering results, the corresponding user groups are divided into related groups to obtain multiple user groups.
[0066] It should be noted that the present invention uses the DBSCAN algorithm to perform association analysis on user characteristics. Through the clustering results, different types of users can be effectively grouped based on dining behavior characteristics, and the user behaviors in the same group have highly similar characteristics. In the clustering results, the imported data in the clustering model is divided into multiple clustering groups. The clustering groups correspond to the user groups one by one.
[0067] According to an embodiment of the present invention, the current user interaction information is obtained based on the target catering platform, user feature analysis and extraction are performed based on the current user interaction information to obtain current user feature data, clustering groups are screened for the current user in combination with the clustering model, and the user group to which the current user belongs is obtained, specifically:
[0068] Analyze real-time user information through the target catering platform, obtain real-time user interaction information and mark it as current user interaction information;
[0069] Perform user feature analysis and feature extraction based on current user interaction information to form current user feature data;
[0070] The current user feature data is imported into the clustering model and the distance to each clustering group is calculated, and the user group to which the current user feature data belongs is determined by the shortest distance.
[0071] It should be noted that the user feature analysis and feature extraction refers to the analysis of the user's behavioral characteristics in catering interaction information, including dimensions such as dish type, ingredient variety, order time, ordering frequency, and consumption amount information.
[0072] According to an embodiment of the present invention, the user feature data in the user group is collected to form a current feature data set, and in the catering dish database, catering data recommendation analysis is performed based on the current feature data set to form a recommendation data set, specifically:
[0073] Through the clustering model, the user feature data within the user group is collected and integrated to form the current feature data set;
[0074] A collaborative filtering recommendation algorithm is introduced, and the current feature data set is used as the user interest feature. The recommendation data is analyzed and screened in the restaurant menu database to obtain the recommendation data set.
[0075] It should be noted that after forming multiple user groups, by analyzing the interactive behavior characteristics of real-time users and screening out the user groups with the closest behaviors, recommendation analysis is performed from the database based on the characteristic data in the user groups to form a recommendation data set. Since the user group generally has a large base, the amount of data obtained from the corresponding recommendation data set is also large. The recommendation data set includes recommendation data of various types and formats, such as dish information, ingredient information, promotion information, etc. The data set has multiple data types and formats such as pictures, videos, audio, web pages, etc., which are suitable for different terminals.
[0076] According to an embodiment of the present invention, the acquisition of display demand information of N preset terminals, the data format condition analysis is performed through the display demand information, and N types of conditional feature data are formed, a classification model based on a decision tree is constructed, and the N types of conditional feature data are used as judgment nodes of the classification model to construct a classification model that forms a complete decision tree, specifically:
[0077] Obtain display demand information of N preset terminals, analyze the adapted data format and data type based on the display demand information, and generate corresponding data condition information;
[0078] Extract feature information based on data condition information to form N types of conditional feature data;
[0079] Build a decision tree-based classification model;
[0080] Based on N kinds of conditional feature data, feature condition conversion based on decision tree nodes is performed to form N conditional nodes;
[0081] Taking N conditional nodes as the initial nodes in the classification model, the classification model is determined based on the heuristic algorithm to determine each node and classification condition, and a classification model based on the complete decision tree is constructed.
[0082] It should be noted that the preset terminals include mobile phones, computers, web pages, tablets, smart wearable devices and other terminals. Various preset terminals are connected to the target catering platform through the Internet and realize data display. The data condition information includes the data type, data format and other related condition setting information of the preset terminal. N types of conditional feature data correspond to N types of preset terminal devices. The heuristic algorithm can adopt algorithms such as ID3, C4.5, and CART.
[0083] For example, in a preferred embodiment, the relevant condition setting information can be different image resolutions, picture formats, video formats, video file sizes, video lengths and other conditions. Different preset terminals correspond to different condition information. The information is converted into conditional features to form conditional branches of a decision tree. By constructing a decision tree model, the recommended data sets can be classified and screened based on different conditions. In addition, the decision tree construction process can be optimized based on a heuristic algorithm, so that the decision tree can realize rapid classification of data sets and realize recommended data display on multiple terminals based on the classified data.
[0084] After the decision tree classification model is built, the decision tree can be trained by a preset test data set. The preset test data set uses the preferred display data that has been stored in each terminal as the test data. The preferred display data of each terminal is collected to form a test data set. Before collection, different display data corresponding to the preset terminal are recorded to record the real classification results. By comparing the predicted classification results of the decision tree with the real results during the test, the branch conditions of the decision tree can be optimized by accuracy. The preferred display data can be based on manual screening.
[0085] In the actual classification process, the recommended data set is used as input data, and through multiple branch conditions, the recommended data set is finally divided into multiple groups of recommended data. Each group of recommended data corresponds to a minimum classification data, and each group of recommended data also corresponds to an applicable one (or more) preset terminal devices, and the classification results are recorded at the same time.
[0086] The main purpose of the classification model is to quickly classify recommended data sets with high data integration, and through the classification results, achieve the technical effect of adapting the classified data to different preset terminals. Compared with the traditional technology of applying the same data set to different preset terminals without screening, the present invention can effectively improve the display effect and user experience of different terminals, so that each terminal can cooperate more stably.
[0087] The classification model can be trained and optimized through the preferred recommendation data of historical terminals. In the classification model, the terminal data format conditions are used as the judgment basis of the node, and the constructed decision tree classification model can realize the rapid classification of the recommendation data set with a large amount of data, form recommendation data suitable for various preset terminals, improve the terminal display effect and the display efficiency of the overall system, reduce the instability between the platform and the terminal, and improve the user experience.
[0088] According to an embodiment of the present invention, the recommended data set is imported into a classification model for data classification, and based on the classification result, the recommended data set is divided into multiple groups of recommended data, and based on the display demand information of N preset terminals, the multiple groups of recommended data are respectively associated with the preset terminals, and a multi-terminal data recommendation scheme is generated based on the association information and the multiple groups of recommended data, specifically:
[0089] Import the recommended data set into the classification model for data classification and obtain the classification results;
[0090] Based on the classification results, the recommendation data set is divided into multiple groups of recommendation data;
[0091] Based on display demand information of N preset terminals, multiple groups of recommended data are associated with the preset terminals respectively, and a multi-terminal data recommendation scheme is generated based on the associated information and the multiple groups of recommended data.
[0092] It should be noted that, in the multi-terminal data recommendation solution based on content demand analysis, different terminals have different recommended display contents. The multi-terminal data recommendation solution can be used for single real-time user analysis and multi-user recommendation analysis.
[0093] According to an embodiment of the present invention, it also includes:
[0094] In one analysis cycle, mark a real-time user, analyze and obtain the multi-terminal data recommendation plan for the real-time user;
[0095] Through K analysis cycles, obtain K multi-terminal data recommendation solutions for real-time users;
[0096] Based on the K-times multi-terminal data recommendation scheme, the recommended data of N preset terminals is collected and recorded to obtain N terminal data sets;
[0097] Based on the real-time user interaction with the target catering platform, the M most frequently used preset terminals are counted and marked as preferred preset terminals;
[0098] Based on the preferred preset terminal, M terminal data sets are obtained from the N terminal data sets, and the M terminal data sets are associated with the M preset terminals to form cache data;
[0099] The cached data is associated with the real-time user by tagging and imported into a system database.
[0100] It should be noted that the analysis cycle is generally the data recommendation cycle corresponding to a user ordering process. K is the number of user preset cycles, and M is the number of preset terminals, which is generally 1 to 2. Each multi-terminal data recommendation scheme corresponds to the data display scheme of N preset terminals. Among the N terminal data sets, each terminal data set is used for one terminal type. The system database is a database for storing cache.
[0101] It is worth mentioning that in the process of user recommendation analysis, since it is necessary to perform user clustering and classification analysis of the corresponding recommendation data set, the data processing volume is large for a catering platform with a large user base. Therefore, the present invention collects data on multiple recommendation plans for users within K analysis cycles, marks the preset terminals used frequently (corresponding to the preferred preset terminals), records the corresponding recommendation data for the high-frequency preset terminals, and forms cache data. The cache data is for the high-frequency preset terminals. Based on the cache data, in subsequent analysis cycles or subsequent real-time user recommendation analysis processes, the real-time users extract the associated data in the corresponding cache data by analyzing them. When the user uses the historical high-frequency preset terminals, the recommendation data can be displayed in real time and quickly, thereby reducing the repeated analysis process of the recommendation data and effectively improving the system efficiency and system stability.
[0102] Figure 2 A block diagram of a multi-terminal meal ordering intelligent management system based on multiple users of the present invention is shown.
[0103] The second aspect of the present invention further provides a multi-terminal meal ordering intelligent management system 2 based on multiple users, the system comprising: a memory 21, a processor 22, the memory comprising a multi-terminal meal ordering intelligent management program based on multiple users, the multi-terminal meal ordering intelligent management program based on multiple users when executed by the processor implements the following steps:
[0104] Through the target catering platform, collect the user ordering interaction data in the historical time period, and perform user feature analysis and feature extraction on multiple users based on the interaction data to form user feature data;
[0105] A clustering model is built based on the DBSCAN algorithm, and user feature data is imported into the clustering model as clustering sample data for clustering grouping. Based on the clustering results, the user group is divided into multiple user groups;
[0106] Based on the target catering platform, the current user interaction information is obtained, and user feature analysis and extraction are performed based on the current user interaction information to obtain the current user feature data, and the clustering group of the current user is screened in combination with the clustering model to obtain the user group to which the current user belongs;
[0107] Collect user feature data within the user group to form a current feature data set, and perform catering data recommendation analysis based on the current feature data set in the catering menu database to form a recommendation data set;
[0108] Obtain display demand information of N preset terminals, perform data format condition analysis based on the display demand information, and form N condition feature data, build a classification model based on a decision tree, use the N condition feature data as judgment nodes of the classification model, and build a classification model that forms a complete decision tree;
[0109] The recommendation data set is imported into the classification model for data classification. Based on the classification results, the recommendation data set is divided into multiple groups of recommendation data. Based on the display demand information of N preset terminals, the multiple groups of recommendation data are respectively associated with the preset terminals, and a multi-terminal data recommendation plan is generated based on the associated information and the multiple groups of recommendation data.
[0110] According to an embodiment of the present invention, the user ordering interaction data in the historical time period is collected through the target catering platform, and user feature analysis and feature extraction are performed on multiple users based on the interaction data to form user feature data, specifically:
[0111] Based on a preset historical time period, the user ordering interaction data of all users is collected through the target catering platform;
[0112] Through the user ordering interaction data, user feature analysis and feature extraction are performed on various users to form user feature data of various users;
[0113] The user characteristic analysis includes dimensions of dish type, ingredient variety, order time, ordering frequency, and consumption amount information.
[0114] It should be noted that the target catering platform includes a user data module, an order module, a database module, a catering interaction module, a display module, etc., which are responsible for different catering information service functions. The historical time period can be the historical operation time period of the platform, or it can be a time period set by the user, specifically a longer time period, which is used to collect enough user data for feature analysis. The user interaction data is the data collected when the user orders food or browses data on various terminals.
[0115] According to an embodiment of the present invention, the clustering model is constructed based on the DBSCAN algorithm, user feature data is imported into the clustering model as clustering sample data for clustering grouping, and the user group is divided into multiple user groups based on the clustering results, specifically:
[0116] Construct a clustering model based on the DBSCAN algorithm;
[0117] The user feature data is associated with the user information, and the user feature data is imported into the clustering model as the clustering sample data. The minimum distance threshold and the minimum number of neighbors are set for data classification and clustering expansion, and finally the clustering results are obtained based on the data set division.
[0118] Based on the clustering results, the corresponding user groups are divided into related groups to obtain multiple user groups.
[0119] It should be noted that the present invention uses the DBSCAN algorithm to perform association analysis on user characteristics. Through the clustering results, different types of users can be effectively grouped based on dining behavior characteristics, and the user behaviors in the same group have highly similar characteristics. In the clustering results, the imported data in the clustering model is divided into multiple clustering groups. The clustering groups correspond to the user groups one by one.
[0120] According to an embodiment of the present invention, the current user interaction information is obtained based on the target catering platform, user feature analysis and extraction are performed based on the current user interaction information to obtain current user feature data, clustering groups are screened for the current user in combination with the clustering model, and the user group to which the current user belongs is obtained, specifically:
[0121] Analyze real-time user information through the target catering platform, obtain real-time user interaction information and mark it as current user interaction information;
[0122] Perform user feature analysis and feature extraction based on current user interaction information to form current user feature data;
[0123] The current user feature data is imported into the clustering model and the distance to each clustering group is calculated, and the user group to which the current user feature data belongs is determined by the shortest distance.
[0124] It should be noted that the user feature analysis and feature extraction refers to the analysis of the user's behavioral characteristics in catering interaction information, including dimensions such as dish type, ingredient variety, order time, ordering frequency, and consumption amount information.
[0125] According to an embodiment of the present invention, the user feature data in the user group is collected to form a current feature data set, and in the catering dish database, catering data recommendation analysis is performed based on the current feature data set to form a recommendation data set, specifically:
[0126] Through the clustering model, the user feature data within the user group is collected and integrated to form the current feature data set;
[0127] A collaborative filtering recommendation algorithm is introduced, and the current feature data set is used as the user interest feature. The recommendation data is analyzed and screened in the restaurant menu database to obtain the recommendation data set.
[0128] It should be noted that after forming multiple user groups, by analyzing the interactive behavior characteristics of real-time users and screening out the user groups with the closest behaviors, recommendation analysis is performed from the database based on the characteristic data in the user groups to form a recommendation data set. Since the user group generally has a large base, the amount of data obtained from the corresponding recommendation data set is also large. The recommendation data set includes recommendation data of various types and formats, such as dish information, ingredient information, promotion information, etc. The data set has multiple data types and formats such as pictures, videos, audio, web pages, etc., which are suitable for different terminals.
[0129] According to an embodiment of the present invention, the acquisition of display demand information of N preset terminals, the data format condition analysis is performed through the display demand information, and N types of conditional feature data are formed, a classification model based on a decision tree is constructed, and the N types of conditional feature data are used as judgment nodes of the classification model to construct a classification model that forms a complete decision tree, specifically:
[0130] Obtain display demand information of N preset terminals, analyze the adapted data format and data type based on the display demand information, and generate corresponding data condition information;
[0131] Extract feature information based on data condition information to form N types of conditional feature data;
[0132] Build a decision tree-based classification model;
[0133] Based on N kinds of conditional feature data, feature condition conversion based on decision tree nodes is performed to form N conditional nodes;
[0134] Taking N conditional nodes as the initial nodes in the classification model, the classification model is determined based on the heuristic algorithm to determine each node and classification condition, and a classification model based on the complete decision tree is constructed.
[0135] It should be noted that the preset terminals include mobile phones, computers, web pages, tablets, smart wearable devices and other terminals. Various preset terminals are connected to the target catering platform through the Internet and realize data display. The data condition information includes the data type, data format and other related condition setting information of the preset terminal. N types of conditional feature data correspond to N types of preset terminal devices. The heuristic algorithm can adopt algorithms such as ID3, C4.5, and CART.
[0136] The classification model can be trained and optimized through the preferred recommendation data of historical terminals. In the classification model, the terminal data format conditions are used as the judgment basis of the node, and the constructed decision tree classification model can realize the rapid classification of the recommendation data set with a large amount of data, form recommendation data suitable for various preset terminals, improve the terminal display effect and the display efficiency of the overall system, reduce the instability between the platform and the terminal, and improve the user experience.
[0137] According to an embodiment of the present invention, the recommended data set is imported into a classification model for data classification, and based on the classification result, the recommended data set is divided into multiple groups of recommended data, and based on the display demand information of N preset terminals, the multiple groups of recommended data are respectively associated with the preset terminals, and a multi-terminal data recommendation scheme is generated based on the association information and the multiple groups of recommended data, specifically:
[0138] Import the recommended data set into the classification model for data classification and obtain the classification results;
[0139] Based on the classification results, the recommendation data set is divided into multiple groups of recommendation data;
[0140] Based on display demand information of N preset terminals, multiple groups of recommended data are associated with the preset terminals respectively, and a multi-terminal data recommendation scheme is generated based on the associated information and the multiple groups of recommended data.
[0141] It should be noted that, in the multi-terminal data recommendation solution based on content demand analysis, different terminals have different recommended display contents. The multi-terminal data recommendation solution can be used for single real-time user analysis and multi-user recommendation analysis.
[0142] The third aspect of the present invention also provides a computer-readable storage medium, which includes an intelligent management program for multi-terminal ordering based on multiple users. When the intelligent management program for multi-terminal ordering based on multiple users is executed by a processor, the steps of the intelligent management method for multi-terminal ordering based on multiple users as described in any one of the above items are implemented.
[0143] The present invention discloses an intelligent management method and system for multi-terminal ordering based on multiple users. The present invention relates to the field of catering recommendation technology. The user ordering interaction data is collected through the target catering platform, and the users are clustered and grouped using the DBSCAN algorithm. Based on the current user interaction information, the user group to which they belong is matched, and the user feature data in the group is collected, and recommendation analysis is performed in combination with the catering menu database. According to the display requirements of N preset terminals, a classification model based on a decision tree is constructed, the recommended data set is classified, and a multi-terminal data recommendation scheme is generated. The present invention realizes accurate and personalized catering data recommendation, improves user satisfaction and recommendation efficiency, and is suitable for dynamic data recommendation under multiple terminals.
[0144] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.
[0145] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.
[0146] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.
[0147] A person of ordinary skill in the art can understand that: all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: a mobile storage device, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.
[0148] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.
[0149] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.
Claims
1. A multi-terminal meal ordering intelligent management method based on multiple users, characterized in that: include: Through the target catering platform, collect the user ordering interaction data in the historical time period, and perform user feature analysis and feature extraction on multiple users based on the interaction data to form user feature data; A clustering model is built based on the DBSCAN algorithm, and user feature data is imported into the clustering model as clustering sample data for clustering grouping. Based on the clustering results, the user group is divided into multiple user groups; Based on the target catering platform, the current user interaction information is obtained, and user feature analysis and extraction are performed based on the current user interaction information to obtain the current user feature data, and the clustering group of the current user is screened in combination with the clustering model to obtain the user group to which the current user belongs; Collect user feature data within the user group to form a current feature data set, and perform catering data recommendation analysis based on the current feature data set in the catering menu database to form a recommendation data set; Obtain display demand information of N preset terminals, perform data format condition analysis based on the display demand information, and form N condition feature data, build a classification model based on a decision tree, use the N condition feature data as judgment nodes of the classification model, and build a classification model that forms a complete decision tree; Import the recommendation data set into the classification model for data classification. Based on the classification results, divide the recommendation data set into multiple groups of recommendation data. Based on the display demand information of N preset terminals, associate the multiple groups of recommendation data with the preset terminals respectively, and generate a multi-terminal data recommendation plan based on the association information and the multiple groups of recommendation data. The acquisition of display demand information of N preset terminals, the data format condition analysis is performed through the display demand information, and N condition feature data are formed, a classification model based on a decision tree is constructed, and the N condition feature data are used as judgment nodes of the classification model to construct a classification model forming a complete decision tree, specifically: Obtain display demand information of N preset terminals, analyze the adapted data format and data type based on the display demand information, and generate corresponding data condition information; Extract feature information based on data condition information to form N types of conditional feature data; Build a decision tree-based classification model; Based on N kinds of conditional feature data, feature condition conversion based on decision tree nodes is performed to form N conditional nodes; Take N conditional nodes as the initial nodes in the classification model, determine each node and classification condition of the classification model based on the heuristic algorithm, and build a classification model based on the complete decision tree; The method of importing the recommended data set into the classification model for data classification, dividing the recommended data set into multiple groups of recommended data based on the classification result, associating the multiple groups of recommended data with the preset terminals based on the display demand information of N preset terminals, and generating a multi-terminal data recommendation scheme based on the association information and the multiple groups of recommended data, is specifically as follows: Import the recommended data set into the classification model for data classification and obtain the classification results; Based on the classification results, the recommendation data set is divided into multiple groups of recommendation data; Based on display demand information of N preset terminals, multiple groups of recommended data are associated with the preset terminals respectively, and a multi-terminal data recommendation scheme is generated based on the associated information and the multiple groups of recommended data.
2. According to claim 1, a method for intelligent management of multi-terminal ordering based on multiple users, characterized in that: The target catering platform collects user ordering interaction data in a historical period, and performs user feature analysis and feature extraction on multiple users based on the interaction data to form user feature data, specifically: Based on a preset historical time period, the user ordering interaction data of all users is collected through the target catering platform; Through the user ordering interaction data, user feature analysis and feature extraction are performed on various users to form user feature data of various users; The user characteristic analysis includes dimensions of dish type, ingredient variety, order time, ordering frequency, and consumption amount information.
3. According to claim 2, a method for intelligent management of multi-terminal meal ordering based on multiple users, characterized in that: The clustering model is constructed based on the DBSCAN algorithm, user feature data is imported into the clustering model as clustering sample data for clustering grouping, and the user group is divided into multiple user groups based on the clustering results, specifically: Construct a clustering model based on the DBSCAN algorithm; The user feature data is associated with the user information, and the user feature data is imported into the clustering model as the clustering sample data. The minimum distance threshold and the minimum number of neighbors are set for data classification and clustering expansion, and finally the clustering results are obtained based on the data set division. Based on the clustering results, the corresponding user groups are divided into related groups to obtain multiple user groups.
4. According to claim 3, a multi-terminal meal ordering intelligent management method based on multiple users is characterized in that: The method of obtaining the current user interaction information based on the target catering platform, performing user feature analysis and extraction based on the current user interaction information, obtaining the current user feature data, screening the clustering group of the current user in combination with the clustering model, and obtaining the user group to which the current user belongs is specifically as follows: Analyze real-time user information through the target catering platform, obtain real-time user interaction information and mark it as current user interaction information; Perform user feature analysis and feature extraction based on current user interaction information to form current user feature data; The current user feature data is imported into the clustering model and the distance to each clustering group is calculated, and the user group to which the current user feature data belongs is determined by the shortest distance.
5. According to claim 4, a method for intelligent management of multi-terminal meal ordering based on multiple users, characterized in that: The user characteristic data in the user group is collected to form a current characteristic data set, and in the catering dish database, catering data recommendation analysis is performed based on the current characteristic data set to form a recommendation data set, specifically: Through the clustering model, the user feature data within the user group is collected and integrated to form the current feature data set; A collaborative filtering recommendation algorithm is introduced, and the current feature data set is used as the user interest feature. The recommendation data is analyzed and screened in the restaurant menu database to obtain the recommendation data set.
6. An intelligent management system for multi-terminal ordering based on multiple users, characterized in that: The system includes: a memory and a processor. The memory includes a multi-terminal meal ordering intelligent management program based on multiple users. When the multi-terminal meal ordering intelligent management program based on multiple users is executed by the processor, the following steps are implemented: Through the target catering platform, collect the user ordering interaction data in the historical time period, and perform user feature analysis and feature extraction on multiple users based on the interaction data to form user feature data; Through the target catering platform, collect the user ordering interaction data in the historical time period, and perform user feature analysis and feature extraction on multiple users based on the interaction data to form user feature data; A clustering model is built based on the DBSCAN algorithm, and user feature data is imported into the clustering model as clustering sample data for clustering grouping. Based on the clustering results, the user group is divided into multiple user groups; Based on the target catering platform, the current user interaction information is obtained, and user feature analysis and extraction are performed based on the current user interaction information to obtain the current user feature data, and the clustering group of the current user is screened in combination with the clustering model to obtain the user group to which the current user belongs; Collect user feature data within the user group to form a current feature data set, and perform catering data recommendation analysis based on the current feature data set in the catering menu database to form a recommendation data set; Obtain display demand information of N preset terminals, perform data format condition analysis based on the display demand information, and form N condition feature data, build a classification model based on a decision tree, use the N condition feature data as judgment nodes of the classification model, and build a classification model that forms a complete decision tree; Import the recommendation data set into the classification model for data classification. Based on the classification results, divide the recommendation data set into multiple groups of recommendation data. Based on the display demand information of N preset terminals, associate the multiple groups of recommendation data with the preset terminals respectively, and generate a multi-terminal data recommendation plan based on the association information and the multiple groups of recommendation data. The acquisition of display demand information of N preset terminals, the data format condition analysis is performed through the display demand information, and N condition feature data are formed, a classification model based on a decision tree is constructed, and the N condition feature data are used as judgment nodes of the classification model to construct a classification model forming a complete decision tree, specifically: Obtain display demand information of N preset terminals, analyze the adapted data format and data type based on the display demand information, and generate corresponding data condition information; Extract feature information based on data condition information to form N types of conditional feature data; Build a decision tree-based classification model; Based on N kinds of conditional feature data, feature condition conversion based on decision tree nodes is performed to form N conditional nodes; Take N conditional nodes as the initial nodes in the classification model, determine each node and classification condition of the classification model based on the heuristic algorithm, and build a classification model based on the complete decision tree; The method of importing the recommended data set into the classification model for data classification, dividing the recommended data set into multiple groups of recommended data based on the classification result, associating the multiple groups of recommended data with the preset terminals based on the display demand information of N preset terminals, and generating a multi-terminal data recommendation scheme based on the association information and the multiple groups of recommended data, is specifically as follows: Import the recommended data set into the classification model for data classification and obtain the classification results; Based on the classification results, the recommendation data set is divided into multiple groups of recommendation data; Based on display demand information of N preset terminals, multiple groups of recommended data are associated with the preset terminals respectively, and a multi-terminal data recommendation scheme is generated based on the associated information and the multiple groups of recommended data.
7. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a multi-terminal ordering intelligent management program based on multiple users. When the multi-terminal ordering intelligent management program based on multiple users is executed by the processor, the steps of the multi-terminal ordering intelligent management method based on multiple users as described in any one of claims 1 to 5 are implemented.
