An operation management method based on AI system
Through the operation and management method based on the AI system, the information push library is dynamically adjusted and optimized, the problem of fixed information push range of the intelligent push platform is solved, personalized and real-time information push is realized, and the user experience is improved.
Patent Information
- Application Number
- CN202310799177.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-06-30
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2043-06-30
AI Technical Summary
The existing intelligent push platform cannot push high-quality information based on users' real-time needs, resulting in a fixed range of information push and cannot meet the changes in users' dynamic needs.
Adopt the operation and management method based on the AI system to initialize the push of popular information, establish a demand model, classify information, collect user feedback, adjust the push library, and realize dynamic adjustment and optimization of information to ensure the personalization and real-timeness of information push.
It realizes dynamic adjustment of the information push library according to user needs and preferences, avoids the problem of fixed information push range, improves the personalization and real-timeness of information push, and improves the user experience.
Smart Images

Figure CN116701768B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of high-quality data operation management, and in particular to an operation management method based on an AI system. Background Art
[0002] Artificial intelligence refers to the use of computer intelligence principles to create computers similar to human brain intelligence, enabling computers to achieve higher-level applications. In today's increasingly innovative science and technology, people are in an information society and need to screen and process all kinds of information to select data that is beneficial to them.
[0003] In the era of big data, information push is indispensable in all aspects of people's lives. At this time, the push platform needs to be able to push high-quality data according to user needs, reduce the push of irrelevant information, and reduce the user experience. The mode of pushing information adopted by the smart push platform on the market is to initialize the information push library first, push popular information to the user end, and then establish an information push library based on the user's search records and user needs and preferences, and push high-quality information that meets the user's search records and user needs and preferences. However, this information push operation and management method has the problem of fixed push information range, and cannot push high-quality information according to the user's real-time needs. For example: after the user retrieves the information, the next time he enters the APP and other smart terminals, he may need not only the same information but also other related and irrelevant information.
[0004] Therefore, we propose an operation management method based on the AI system. Summary of the invention
[0005] The purpose of the present invention is to solve the problem of fixed push information range in the intelligent push platform in the prior art, and to propose an operation management method based on the AI system.
[0006] In order to achieve the above object, the present invention adopts the following technical solutions:
[0007] An operation management method based on an AI system includes the following specific information push operation management steps:
[0008] S1: Initialize push, push popular information to the user end, and establish an information push library;
[0009] S2: Establish a demand model, build a model for user demand information through login information, publicly available information and user search information, and optimize the information push library;
[0010] S3: Information classification, classify and number the information pushed to the user end;
[0011] S4: Preference feedback: after pushing information to the user end, the user's demand for information and preference level are fed back to the push intelligent end to determine the user's information needs and further optimize the information push library;
[0012] S5: Push library adjustment, adjusting the demand range in the push library according to the user's retrieval demand information;
[0013] S6: Ordered push, delivering a small proportion of filtered disordered information and a large proportion of demand and preference information to the user end.
[0014] Preferably, in step S1, the information push library is a popular information push library, and the popular information includes information with a high search rate, information with a high click rate, and information with a high preference rate.
[0015] Preferably, in step S2, the user's login information includes basic information such as gender, age, education level and occupation, and the publicly disclosed information includes information on preferences and demands for interests and hobbies when registering the smart push terminal. Optimizing the information push library refers to searching a large database for information that has a high degree of overlap with the user's basic information and preference demand information.
[0016] Preferably, in step S3, the push information is classified according to user needs and preferences, the demand information refers to the retrieval demand information, the demand information is classified into a first category of push information, the approved push information is classified into a second category of push information, and the unapproved push information is classified into a third category of push information.
[0017] Preferably, when the push information is classified according to the information content, the first category push information, the second category push information and the third category push information are refined according to the specific display content of the information content.
[0018] Preferably, in step S4, when the user actually uses the intelligent push platform, data on user search and whether the push information is accepted are recorded, and an information classification library is established according to user needs and preferences.
[0019] Preferably, in step S5, when the user terminal's search requirement information is within the three types of push information, the range of the three types of push information is narrowed.
[0020] Preferably, in step S6, the unordered information is information other than the first, second and third push information in the information push library, and a large proportion of the demand and preference information is the first and second push information in the information push library.
[0021] Compared with the prior art, the present invention has the following beneficial effects:
[0022] 1. After being divided according to user needs and preferences, the information under the three classified information can be adjusted according to user search needs. When the search need information is in the second category of push information, it is promoted to the first category of push information. When the search need information is in the third category of push information, it is promoted to the second category of push information. When the search need information is in disordered information, the information about the search need information in the disordered information is classified into the first category of push information. In this way, the information push library can be freely adjusted to avoid the fixed range of information pushed by the intelligent push platform, thereby solving the problem of fixed range of information pushed by the intelligent push platform in the prior art mentioned in the above background technology.
[0023] 2. After classification according to user needs and preferences, each type of information is divided into Class I push information, Class II push information, Class III push information and unordered information. The subsequent intelligent push information is pushed to the user end in different proportions based on the classification, so as to achieve the effect of intelligent information push, so that only information data that meets user needs and preferences will be pushed during the push. BRIEF DESCRIPTION OF THE DRAWINGS
[0024] Figure 1 This is a flowchart of the information push operation management steps of an operation management method based on an AI system proposed by the present invention; DETAILED DESCRIPTION
[0025] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.
[0026] Reference Figure 1 , an operation management method based on an AI system, including the following specific information push operation management steps:
[0027] S1: Initialize push, push popular information to the user end, and establish an information push library;
[0028] S2: Establish a demand model, build a model for user demand information through login information, publicly available information and user search information, and optimize the information push library;
[0029] S3: Information classification, classify and number the information pushed to the user end;
[0030] S4: Preference feedback: after pushing information to the user end, the user's demand for information and preference level are fed back to the push intelligent end to determine the user's information needs and further optimize the information push library;
[0031] S5: Push library adjustment, adjusting the demand range in the push library according to the user's retrieval demand information;
[0032] S6: Ordered push, delivering a small proportion of filtered disordered information and a large proportion of demand and preference information to the user end.
[0033] Through the above technical solution, in step S3, the push information is classified according to the user's needs and preferences. The demand information refers to the retrieval demand information, and the demand information is classified into a first category of push information, the approved push information is classified into a second category of push information, and the unapproved push information is classified into a third category of push information. When the push information is classified according to the information content, the first category of push information, the second category of push information, and the third category of push information are refined according to the specific display content of the information content;
[0034] Based on the above, when classifying and numbering according to the content of information, it can be further refined according to the push content of the smart push platform. For example, when pushing news, the pushed news information can be divided into: political news, economic news, legal news, military news, science and technology news, cultural and educational news, sports news, social news, etc., and the specific pushed content can be numbered according to the English letters. When classified according to user needs and preferences, the first category of push information is numbered "6", the second category of push information is numbered "3", and the third category of push information is numbered "0", among which the disordered information that is not in the three major categories is numbered "1".
[0035] Specifically, in step S1, the information push library is a popular information push library, and the popular information includes information with high retrieval rate, information with high click rate and information with high preference rate. In step S2, the user's login information includes basic information such as gender, age, education level and occupation, and the publicly available information includes information on interests and hobbies when registering for the smart push terminal. Optimizing the information push library refers to retrieving information with a high degree of overlap with the user's basic information and preference information by searching a large database.
[0036] Through the above technical solution, the popular information push library is established in the middle and late stages of the operation of the smart push terminal. In the initial stage, the basic information of the user's login, interests and hobbies are used as the framework to push information that meets the user's basic information and preference information, such as: pushing information that meets the user's age group through age information. This push operation management method is commonly seen in video push APPs.
[0037] Specifically, in step S4, when the user actually uses the intelligent push platform, data on the user's search and whether the push information is accepted are recorded, and an information classification library is established according to user needs and preferences.
[0038] Through the above technical solution, after classification in step S3, an information classification number that meets the user's needs and preferences is established according to the user's specific search content and attitude towards the pushed information.
[0039] Specifically, in step S5, when the user terminal retrieves the demand information within the three categories of push information, the scope of the three categories of push information is narrowed. In step S6, the unordered information is information other than the first category push information, the second category push information and the third category push information in the information push library. A large proportion of the demand and preference information is the first category push information and the second category push information in the information push library.
[0040] Through the above technical solution, the push ratio of each type of information in step S6 can be obtained by multiplying the number of each type of push information by 0.1 when classifying according to user needs and preferences, and the push ratio of each type of push information pushed to the user terminal at the smart push terminal;
[0041] Based on the above, in step S5, this type of situation is common when the user retrieves work demand content. In the leisure state, the work demand content is the third category of push information, but when working again, the work demand content is retrieved as information classified under the third category of push information. Through the user's active retrieval, the work demand content is promoted from the third category of push information to the second category of push information, avoiding direct promotion to the first category of push information, so that a large amount of work demand content can be retrieved during the leisure stage.
[0042] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.
Claims
1. An operation management method based on an AI system, characterized in that: The following specific information push operation management steps are included: S1: Initialize push, push popular information to the user end, and establish an information push library; S2: Establish a demand model, build a model for user demand information through login information, publicly available information and user search information, and optimize the information push library; S3: Information classification, classify and number the information pushed to the user end; S4: Preference feedback: after pushing information to the user end, the user's demand for information and preference level are fed back to the push intelligent end to determine the user's information needs and further optimize the information push library; S5: Push library adjustment, adjusting the demand range in the push library according to the user's retrieval demand information; S6: Ordered push, delivering a small proportion of filtered disordered information and a large proportion of demand and preference information to the user end; In step S3, the push information is classified according to the user's needs and preferences. The demand information refers to the retrieval demand information. The demand information is classified into the first category of push information, the approved push information is classified into the second category of push information, and the unapproved push information is classified into the third category of push information; In step S4, when the user actually uses the intelligent push platform, the data of the user's search and whether the push information is accepted are recorded, and an information classification library is established according to the user's needs and preferences; In step S5, when the user terminal searches for the required information within the three types of push information, the range of the three types of push information is narrowed; In step S6, the unordered information is information other than the first, second and third push information in the information push library, and a large proportion of the demand and preference information is the first and second push information in the information push library.
2. The operation management method based on AI system according to claim 1, characterized in that: In step S1, the information push library is a popular information push library, and the popular information includes information with high retrieval rate, information with high click rate and information with high preference rate.
3. The operation management method based on AI system according to claim 1, characterized in that: In step S2, the user's login information includes basic information such as gender, age, education level and occupation. The publicly available information includes information on interests and hobbies when registering for the smart push terminal. Optimizing the information push library refers to searching a large database for information that has a high degree of overlap with the user's basic information and preference information.
4. The operation management method based on AI system according to claim 1, characterized in that: When push information is classified according to information content, the first category push information, the second category push information and the third category push information are refined according to the specific display content of the information.
