Knowledge database updating method for AI agent

Abstract

The invention relates to the technical field of data processing, in particular to a knowledge database updating method for an AI agent, and the method comprises the steps: when a user uses the AI agent, obtaining updating data of each time and each keyword of the updating data of each time; the historical difference degree of each keyword of the data is updated at the current time; screening each candidate keyword; obtaining a unique feature value of the keyword of the current updated data after each candidate keyword is removed; obtaining the keyword uniqueness of the current updated data; extracting each feature keyword; and on the basis of the feature keyword, obtaining a unique identifier of the current updated data, storing the unique identifier into a knowledge database of the AI agent, and searching and updating the data through the unique identifier. The method and the device aim at improving the updating efficiency and the management efficiency of the knowledge database by reducing the calculation time consumption of the unique identifier and improving the uniqueness of the unique identifier.

Description

Technical Field

[0001] This application relates to the technical field of data processing, and particularly to a method for updating a knowledge database for an AI agent. Background Art

[0002] An AI agent is an intelligent system that can perceive the environment and achieve specific goals through autonomous decision-making. Its core part is a knowledge database, which is used to store and manage various information, including facts, rules, experiences, etc., and provide a basis for the decision-making of the AI agent. The update method of the knowledge database is crucial for the performance of the AI agent. Incremental update is an efficient update strategy that allows the AI agent to gradually add, modify, or delete information based on the original knowledge, rather than reconstructing the entire database every time there is an update. This method can improve the update efficiency, reduce the consumption of computing resources, ensure the continuity and consistency of knowledge, and at the same time enable the AI agent to quickly adapt to environmental changes and integrate new knowledge in a timely manner, thereby enhancing the adaptability and flexibility of the AI agent.

[0003] In the prior art for incremental updates, the content hashing method is usually adopted, and data is called by calculating the hash value of each record or file. However, due to the large amount of data stored in the AI agent and the relatively long length of each piece of data, the complexity of hash calculation is closely related to the data length. The longer the data, the more resources are required for calculation. When the prior art directly performs hash calculation on the original data, there is a defect of long time consumption, resulting in waste of computer resources and low efficiency. Summary of the Invention

[0004] In view of the above, it is necessary to provide a method for updating a knowledge database for an AI agent, which can improve the update efficiency and management efficiency of the knowledge database by reducing the calculation time consumption of the unique identifier and improving the uniqueness of the unique identifier compared with the traditional method for updating the knowledge database for an AI agent.

[0005] A method for updating a knowledge database for an AI agent according to this application adopts the following technical solution:

[0006] An embodiment of this application provides a method for updating a knowledge database for an AI agent, and this method includes the following steps:

[0007] When a user uses the AI agent, obtain the update data each time and each keyword of the update data each time;

[0008] Obtain the historical difference degree of each keyword by the proportion of the number of times that each keyword in the current update data does not appear in the remaining update data, and combine the difference in the occurrence frequency of each keyword in the current update data and the remaining update data;

[0009] Screen each candidate keyword from the keywords of the current update data according to the historical difference degree; sequentially remove each candidate keyword from all candidate keywords to obtain each candidate keyword set, and obtain the number of elements that do not appear in the remaining update data each time for all elements in each candidate keyword set, and compare the occurrence frequencies of each element in the current update data and the remaining update data each time to obtain the uniqueness feature value of the keywords of the current update data after removing each candidate keyword;

[0010] Obtain the keyword uniqueness of the current update data through the number of occurrences that all candidate keywords obtained by screening do not appear in the remaining update data each time, and the difference in the occurrence frequencies of each candidate keyword between the current update data and the remaining update data each time;

[0011] Extract each feature keyword from the screened candidate keywords through the difference between the uniqueness feature value and the keyword uniqueness; based on the feature keyword, obtain the unique identifier of the current update data and store it in the knowledge database of the AI intelligent agent, and search for and update the data through the unique identifier.

[0012] In one embodiment, the process of obtaining the historical difference degree is as follows:

[0013] Calculate the difference amount between the occurrence frequencies of each keyword in the current update data and the remaining update data each time;

[0014] Calculate the weighted sum value of the difference amounts between each keyword in the current update data and all the remaining update data, where the weights of all the difference amounts are equal and the sum of the weights is 1;

[0015] The historical difference degree can be further obtained through the ratio of the number of times and the weighted sum value, where the ratio of the number of times is not 0.

[0016] In one embodiment, the historical difference degree is the product of the ratio of the number of times and the weighted sum value.

[0017] In one embodiment, the method for screening each candidate keyword from the keywords of the current update data is as follows:

[0018] According to the magnitude of the historical difference degree of all keywords in the current update data, divide all keywords into two categories, and use each keyword in the category with the largest historical difference degree as each candidate keyword.

[0019] In one embodiment, the process of obtaining the uniqueness feature value is as follows:

[0020] Calculate the cumulative sum of the number of elements and a preset positive integer;

[0021] Denote the difference in the frequency of occurrence of each element in each candidate keyword set between the current update data and each of the remaining update data as the frequency difference; calculate the mean of all positive values among the frequency differences of all elements in each candidate keyword set between the current update data and each of the remaining update data.

[0022] The uniqueness eigenvalue can be further obtained from the sum of the remaining all update data and the mean value.

[0023] In one embodiment, the calculation method of the uniqueness eigenvalue is as follows:

[0024] Calculate the product of the sum of the remaining all update data and the mean value.

[0025] The uniqueness eigenvalue is the sum value of the products of the remaining all update data.

[0026] In one embodiment, the process of obtaining the keyword uniqueness is as follows:

[0027] Calculate the cumulative value of the number and a preset positive integer.

[0028] Calculate the difference calculation result of the frequency of occurrence of each candidate keyword between the current update data and each of the remaining update data; calculate the average value of the positive values among the difference calculation results corresponding to all the candidate keywords obtained by screening.

[0029] Calculate the product calculation result of the cumulative value of the remaining all update data and the average value.

[0030] The keyword uniqueness is the sum value of the product calculation results of the remaining all update data.

[0031] In one embodiment, the extraction method of the feature keyword is: Denote the difference between the uniqueness eigenvalue and the keyword uniqueness as the unique difference, and extract each feature keyword from the candidate keywords obtained by screening according to the magnitude of the unique difference.

[0032] In one embodiment, extracting each feature keyword from the candidate keywords obtained by screening includes: When the unique difference is less than or equal to 0, the corresponding candidate keywords to be excluded are used as feature keywords.

[0033] In one embodiment, the acquisition method of the unique identifier is as follows:

[0034] Arrange the feature keywords and the occurrence frequencies of the feature keywords in the current updated data in descending order of occurrence frequency to obtain a feature keyword string, add the update date of the current updated data at the end of the feature keyword string to obtain a unique feature string, and combine it with a hashing algorithm to obtain a unique identifier.

[0035] The present application has at least the following beneficial effects:

[0036] By calculating the historical difference degree, the present application can quantify the difference degree of keywords between the current updated data and the historical data, helping to identify keywords with significant changes; through the historical difference degree, candidate keywords with representativeness and distinctiveness are screened out for subsequent extraction of feature keywords to generate a unique identifier, avoiding the problem that directly using all keywords may lead to overly complex generation of unique identifiers and difficulty in ensuring uniqueness; further, by eliminating the uniqueness differences of the keywords in the current updated data before and after each candidate keyword, the distinctiveness of each candidate keyword is evaluated, providing a more accurate basis for subsequent generation of a unique identifier; by extracting the most representative feature keywords, a unique identifier for the current updated data is generated, reducing the time-consuming of hash calculation. At the same time, the generated unique identifier can accurately reflect the core features of the current updated data, effectively avoiding data duplication and confusion, ensuring the uniqueness of each updated data in the knowledge database, and being able to quickly locate and update data when updating the data in the knowledge database, improving the update efficiency and management efficiency of the knowledge database. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0038] Figure 1 It is a flowchart of the steps of a method for updating a knowledge database for an AI intelligent agent provided by the present application;

[0039] [[ID=’18]] Figure 2 It is a schematic diagram of the process for obtaining the keyword distinctiveness; [[ID=‘20]]

[0040] Figure 3 It is a schematic diagram of the process for extracting feature keywords. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0041] In the description of the embodiments of the present application, words such as "exemplary", "or", "for example", etc. are used to represent examples, illustrations or explanations. Any embodiment or design solution described as "exemplary" or "for example" in the embodiments of the present application should not be construed as being more preferred or more advantageous than other embodiments or design solutions. Rather, the use of words such as "exemplary", "or", "for example" is intended to present relevant concepts in a specific manner.

[0042] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which this application belongs. The terms used in this application are only for the purpose of describing specific embodiments and are not intended to limit this application. It should be understood that unless otherwise specified in this application, " / " means "or".

[0043] In addition, it should be noted that the terms "first" and "second" in this application are used to distinguish similar objects and are not used to describe a specific order or sequence.

[0044] The following specifically describes the specific solution of a method for updating a knowledge database for an AI agent provided by this application with reference to the accompanying drawings.

[0045] A method for updating a knowledge database for an AI agent provided by an embodiment of this application. Specifically, a method for updating a knowledge database for an AI agent is provided as follows. Please refer to Figure 1 This method includes the following steps:

[0046] Step 1, when the user uses the AI agent, obtain the update data each time and each keyword of the update data each time.

[0047] When the user uses the AI agent, files and data will be input to the AI agent. The usage data of the user each time is obtained, and the obtained usage data of the user is preprocessed using regular expressions to remove meaningless characters such as ".", ",", "of", etc.; then, the jieba.cut() function of the python database is used to perform word segmentation on the preprocessed data; after word segmentation is completed, keywords are extracted through the extract_tags() function; finally, the update data each time and each keyword of the update data each time are obtained. Among them, regular expressions, the jieba.cut() function, and the extract_tags() function are all well-known technologies and will not be elaborated in this application.

[0048] Step 2, obtain the historical difference degree of each keyword by calculating the proportion of the number of times each keyword in the current update data does not appear in the update data of the remaining times, and combining the frequency difference of each keyword in the current update data and the update data of the remaining times.

[0049] When performing an incremental update operation on a database, a unique identifier is usually used to accurately locate the data or files to be updated. With the precise guidance of the unique identifier, the update process only modifies the target data, strictly ensuring that other data contents are not interfered with and the data sorting remains unchanged. This method effectively avoids redundant scanning and processing of the entire database, thus significantly reducing the computer operation time, improving the data update efficiency, and at the same time ensuring the integrity and consistency of the data, providing strong support for the efficient management and maintenance of the database.

[0050] Among them, the unique identifier is unique. Since the AI agent database contains a large amount of data, when extracting unique features, due to the complexity and diversity of the data, as well as the potential uncertainties in the data generation and storage processes, the unique identifier may be repeated during the extraction process, resulting in different data items being assigned the same unique feature, thereby causing problems such as data confusion or update errors.

[0051] Based on the above analysis, count the occurrence frequencies of each keyword in the current update data in each of the remaining update data, and form a keyword frequency matrix to represent the repeatability of the keywords and the number of times they appear repeatedly. Each row in the keyword frequency matrix represents the occurrence frequency of a keyword in all the remaining update data. Arrange the occurrence frequencies of each keyword in different update data in the order of the implementation sequence. Each column in the keyword frequency matrix represents the occurrence frequencies of all the keywords in the current update data in each of the remaining update data.

[0052] Furthermore, obtain the historical difference degree of each keyword in the current update data by calculating the proportion of the number of times each keyword in the current update data does not appear in the remaining update data and combining the difference in the occurrence frequencies of each keyword in the current update data and the remaining update data. The expression is:

[0053] ; In the formula, represents the historical difference degree of the keyword corresponding to the i-th row of the keyword frequency matrix in the current update data; represents the number of zero elements in the i-th row of the keyword frequency matrix; a represents a preset positive integer used to avoid the numerator being 0, resulting in the historical difference degree of the keyword being directly 0. The value of a is preset manually, and the implementer can set it by himself. In this embodiment, the value of a is 1; n represents the number of columns of the keyword frequency matrix; represents the element in the i-th row and j-th column of the keyword frequency matrix; represents the occurrence frequency of the keyword corresponding to the i-th row of the keyword frequency matrix in the current update data in the current update data; represents the absolute value operation.

[0054] It should be noted that: the greater the difference in the occurrence frequency of any keyword in the current update data compared to its occurrence frequency in the historical update data, the higher the uniqueness of the said any keyword; at the same time, if the number of times any keyword does not appear in the historical update data is more, that is, it appears less or never in the historical data, then the uniqueness of the said any keyword is higher; in the above cases, the historical difference degree of the keyword will be greater, indicating that the said any keyword is more distinguishable in the current update. By optimizing the incremental update strategy through the said any keyword, the uniqueness of the data can be ensured and the management efficiency can be improved.

[0055] Step 3, through the historical difference degree, screen each candidate keyword from the keywords of the current update data; sequentially eliminate each candidate keyword from all candidate keywords to obtain each candidate keyword set. By the number of elements that do not appear in the remaining each update data in all elements of each candidate keyword set, and comparing the occurrence frequencies of each element in the current update data and the remaining each update data, obtain the uniqueness characteristic value of the keywords of the current update data after eliminating the said each candidate keyword.

[0056] When selecting keywords to calculate the unique identifier of the update data, keywords with a large difference from the historical update data need to be selected. Therefore, according to the magnitude of the historical difference degree of all keywords in the current update data, all keywords are divided into two categories, and each keyword in the category with the largest historical difference degree is used as each candidate keyword. Among them, when dividing all keywords into two categories, algorithms that can be used include the mean division method, the threshold division algorithm, the clustering algorithm, etc. In this embodiment, the mean division method is adopted, specifically: calculate the mean value of the historical difference degree of all keywords in the current update data, and form one category with all keywords in the current update data whose historical difference degree is greater than or equal to the mean value, and form another category with all keywords in the current update data whose historical difference degree is less than the mean value. Arrange the row elements corresponding to all the selected candidate keywords in the keyword frequency matrix in the original order to form a characteristic frequency matrix.

[0057] Form a frequency column vector with the occurrence frequencies of all the selected keywords in the current update data, where the frequency column vector corresponds to the same candidate keyword as each same row of the characteristic frequency matrix. Subtract the elements of each column in the characteristic frequency matrix from the frequency column vector respectively to obtain a frequency difference column vector for each column, and form a frequency difference matrix, which is used to characterize the keyword information difference degree between the current update data and the remaining update data. Denote the elements in the frequency difference matrix as frequency differences.

[0058] In addition, since a certain keyword may affect the degree of difference between the current update data and the data of the remaining updates, it is necessary to calculate the uniqueness eigenvalue of the keywords in the current update data after removing the certain keyword. The expression is as follows:

[0059] ; where represents the uniqueness eigenvalue of the keywords in the current update data after removing the k-th candidate keyword; m represents the number of times of the remaining update data; represents the number of 0 elements in the j-th column element after removing the corresponding row of the k-th candidate keyword in the feature frequency matrix; b represents a preset positive integer used to avoid being 0. The value of b is preset manually, and the implementer can set it by himself. In this embodiment, the value of b is 1; represents the average value of the positive values in the j-th column element after removing the corresponding row of the k-th candidate keyword in the frequency difference matrix.

[0060] It should be noted that: after removing the k-th candidate keyword from the current update data, the frequency change of the remaining candidate keywords can reflect the contribution degree of the k-th candidate keyword to the overall uniqueness. If the uniqueness eigenvalue of the keywords in the current update data is smaller after removing the k-th candidate keyword, it indicates that the k-th candidate keyword plays a more important role in maintaining the uniqueness of the data. After removing the k-th candidate keyword, all the remaining candidate keywords form a candidate keyword set.

[0061] Step 4: Obtain the keyword uniqueness of the current update data by screening the number of times that all the candidate keywords obtained do not appear in each of the remaining update data, and the difference in the appearance frequency between each candidate keyword in the current update data and each of the remaining update data.

[0062] Furthermore, the keyword uniqueness of the current update data is obtained by screening the number of times that all the candidate keywords obtained do not appear in each of the remaining update data, and the difference in the appearance frequency between each candidate keyword in the current update data and each of the remaining update data. The expression is as follows:

[0063] ; where represents the keyword uniqueness of the current update data; m represents the number of times of the remaining update data; represents the number of 0 elements in the j-th column element of the feature frequency matrix; represents the average value of the positive values in the j-th column element of the frequency difference matrix; f represents a preset positive integer used to avoid being 0. The value of f is preset manually, and the implementer can set it by himself. In this embodiment, the value of f is 1.

[0064] It should be noted that: the greater the difference between the current update data and the data of other updates, the greater the difference in the frequency of occurrence of keywords. At the same time, when a certain keyword does not appear in the data of other updates, the uniqueness of the certain keyword is higher. Overall, the uniqueness of the keywords in the current update data increases, and the unique identifier calculated using these keywords is more unique, and the similarity with the unique identifiers of other update data is lower. The schematic diagram of the process for obtaining the keyword uniqueness is as shown in Figure 2 shown below.

[0065] Step 5: Extract each characteristic keyword from the selected candidate keywords based on the difference between the uniqueness eigenvalue and the keyword uniqueness; based on the characteristic keywords, obtain the unique identifier of the current update data, store it in the knowledge database of the AI intelligent agent, and search for and update the data through the unique identifier.

[0066] After removing each candidate keyword, the difference between the uniqueness eigenvalue of the keywords in the current update data and the keyword uniqueness of the current update data is denoted as the unique difference. When the unique difference is less than or equal to 0, the corresponding removed candidate keyword is used as the characteristic keyword. Arrange all the characteristic keywords and their frequencies of occurrence in the current update data in descending order of the frequency of occurrence to obtain the characteristic keyword string. For example: the characteristic keywords are: "management", "production", "output", "date", and their frequencies of occurrence in the current update data are 4, 35, 3, and 21 respectively. Then the characteristic keyword string is: "production 35 date 21 management 4 output 3". Further, to avoid duplication of the characteristic keyword string, considering that time is unique, the update date is added to the characteristic keyword string to obtain the characteristic unique string. For example, if the time is 12:08 on August 10, 2024, the characteristic unique string is "production 35 date 21 management 4 output 3202408101208". The schematic diagram of the process for extracting characteristic keywords is as shown in Figure 3 shown below.

[0067] Furthermore, use the characteristic unique string as the input of the hash algorithm, and the output is the unique identifier. Then, store the current update data and its unique identifier in the knowledge database of the AI intelligent agent to achieve incremental update of the data. When updating the data stored in the knowledge database, identify the data by searching for the unique identifier of the data, obtain the storage location of the data in the knowledge database through the unique identifier, and then add, delete, or modify the data. Among them, the calculation process of the hash algorithm is well-known technology and will not be elaborated in this application.

[0068] In summary, by calculating the historical difference degree, the present application can quantify the difference degree of keywords in the current updated data and historical data, helping to identify keywords with significant changes; through the historical difference degree, representative and distinguishable candidate keywords are screened to be used for subsequently extracting feature keywords from them to generate a unique identifier, avoiding the problem that directly using all keywords may lead to the generation of the unique identifier being too complex and difficult to ensure uniqueness; further, by eliminating the uniqueness difference of the keywords in the current updated data before and after each candidate keyword, the uniqueness degree of each candidate keyword is evaluated, providing a more accurate basis for subsequently generating the unique identifier; by extracting the most representative feature keywords, a unique identifier for the current updated data is generated, reducing the time consumption of hash calculation. At the same time, the generated unique identifier can accurately reflect the core features of the current updated data, effectively avoiding data duplication and confusion, ensuring the uniqueness of each updated data in the knowledge database, and being able to quickly locate and update data when updating the data in the knowledge database, improving the update efficiency and management efficiency of the knowledge database.

[0069] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a segment of a program, or a part of code, and the module, the segment of the program, or the part of code contains one or more executable instructions for implementing the specified logical function. In some alternative implementations, the functions marked in the block may occur in a different order from that marked in the accompanying drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, which may depend on the functions involved. In the descriptions corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in a different order from that disclosed in the description. Sometimes, there is no specific order between different operations or steps. For example, two consecutive operations or steps may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, which may depend on the functions involved. Each block in the block diagram and / or flowchart, as well as the combination of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0070] For those skilled in the art, it is obvious that the present application is not limited to the details of the above exemplary embodiments, and without departing from the basic features of the present application, the present application can be implemented in other specific forms. Therefore, from any point of view, the above embodiments of the present application should be regarded as exemplary and non-restrictive.

Claims

1. A method for updating a knowledge database for an AI agent, characterized in that, The method includes the following steps: When the user uses the AI agent, obtain the update data each time and each keyword of each update data; Obtain the historical difference degree of each keyword by the proportion of the number of times that each keyword in the current update data does not appear in the remaining update data, and combine the frequency difference of each keyword in the current update data and the remaining update data; Through the historical difference degree, screen each candidate keyword from the keywords of the current update data; sequentially eliminate each candidate keyword from all candidate keywords to obtain each candidate keyword set. By the number of elements that do not appear in the remaining update data each time in each candidate keyword set, and compare the appearance frequencies of each element in the current update data and the remaining update data each time, obtain the uniqueness characteristic value of the keywords of the current update data after eliminating each candidate keyword; Obtain the keyword uniqueness degree of the current update data by the number of times that all candidate keywords obtained by screening do not appear in the remaining update data each time, and the difference in the appearance frequencies of each candidate keyword between the current update data and the remaining update data each time; Extract each characteristic keyword from the screened candidate keywords through the difference between the uniqueness characteristic value and the keyword uniqueness degree; based on the characteristic keywords, obtain the unique identifier of the current update data, and store it in the knowledge database of the AI agent, and search for and update the data through the unique identifier.

2. The knowledge database update method for an AI agent according to claim 1, characterized in that, The process of obtaining the historical difference degree is as follows: Calculate the difference amount between the appearance frequencies of each keyword in the current update data and the remaining update data each time; Calculate the weighted sum value of the difference amounts of each keyword between the current update data and all the remaining update data, where the weights of all the difference amounts are equal and the sum of the weights is 1; The historical difference degree can be further obtained through the proportion of the number of times and the weighted sum value, where the proportion of the number of times is not 0.

3. A method for updating a knowledge database for an AI agent according to claim 2, characterized in that, The historical difference degree is the product of the proportion of the number of times and the weighted sum value.

4. A method for updating a knowledge database for an AI agent according to claim 1, characterized in that, The method of screening each candidate keyword from the keywords of the current update data is as follows: According to the magnitude of the historical difference degree of all keywords in the current update data, divide all keywords into two categories, and use the keywords in the category with the largest historical difference degree as each candidate keyword.

5. A method for updating a knowledge database for an AI agent as claimed in claim 1, characterized in that, The process of obtaining the uniqueness characteristic value is as follows: Calculate the cumulative sum of the number of elements and a preset positive integer; Record the difference in the appearance frequencies of each element in each candidate keyword set between the current update data and the remaining update data each time as the frequency difference; calculate the mean value of all positive values in the frequency differences of all elements in each candidate keyword set between the current update data and the remaining update data each time; The uniqueness characteristic value can be further obtained through the cumulative sum of the remaining all update data and the mean value.

6. The knowledge database update method for an AI agent according to claim 5, wherein, The calculation method of the uniqueness characteristic value is as follows: Calculate the product of the cumulative sum of the remaining update data each time and the mean value; The uniqueness characteristic value is the sum value of the products of the remaining all update data.

7. A method for updating a knowledge database for an AI agent according to claim 1, characterized in that, The process of obtaining the keyword uniqueness degree is as follows: Calculate the cumulative value of the number and a preset positive integer; Calculate the difference calculation results of the occurrence frequencies of each candidate keyword between the current updated data and each of the remaining updated data; calculate the average value of the positive values in the difference calculation results of all the candidate keywords obtained by screening between the current updated data and each of the remaining updated data; Calculate the product calculation results of the cumulative value of each of the remaining updated data and the average value; The keyword uniqueness is the sum value of the product calculation results of all the remaining updated data.

8. The knowledge database updating method for an AI agent according to claim 1, wherein The extraction method of the feature keyword is: Denote the difference between the uniqueness feature value and the keyword uniqueness as the unique difference, and extract each feature keyword from the screened candidate keywords according to the magnitude of the unique difference.

9. A method for updating a knowledge database for an AI agent according to claim 8, characterized in that, The extraction of each feature keyword from the screened candidate keywords includes: When the unique difference is less than or equal to 0, the corresponding candidate keyword to be excluded is used as the feature keyword.

10. A method for updating a knowledge database for an AI agent as claimed in claim 1, characterized in that, The acquisition method of the unique identifier is: Arrange the feature keyword and the occurrence frequency of the feature keyword in the current updated data in descending order of the occurrence frequency to obtain a feature keyword string, add the update date of the current updated data at the end of the feature keyword string to obtain a feature unique string, and combine with a hashing algorithm to obtain a unique identifier.

Images