Knowledge base processing method and system based on Internet of Things big data

Abstract

The invention relates to the technical field of knowledge bases, and discloses a knowledge base processing method and system based on Internet of Things big data, and the method comprises the steps: constructing an initial knowledge base, collecting use records, generating a real-time knowledge generation strategy, generating real-time knowledge, and updating the knowledge base. The system corresponds to the method. According to the method, the real-time knowledge generation strategy is obtained by collecting the use record of the user for the initial knowledge base in real time, analyzing the characteristics of the use record and adjusting the initial knowledge generation strategy, so that the knowledge generation can meet the actual demand of the user, and the real-time knowledge generation strategy is utilized to process the real-time Internet of Things big data to generate the real-time knowledge and update the initial knowledge base. According to the method, dynamic updating of the knowledge base is achieved, timeliness and accuracy of the knowledge base are guaranteed, the problems that the knowledge base is not comprehensive in construction, not timely in updating and difficult to adapt to big data characteristics of the Internet of Things are effectively solved, and more accurate and effective knowledge support is provided for application of the Internet of Things.

Description

Technical Field

[0001] The present application relates to the technical field of knowledge bases, and specifically to a knowledge base processing method and system based on Internet of Things big data. Background Art

[0002] At present, with the rapid development of the Internet of Things, the full utilization of Internet of Things big data can bring great value. Specifically, the knowledge base constructed based on Internet of Things big data can provide strong support for various fields. However, the inventor has found the following problems in the existing knowledge bases constructed based on Internet of Things big data: On the one hand, when traditional knowledge base construction methods process Internet of Things big data, it is difficult to effectively integrate massive and heterogeneous data, resulting in incomplete knowledge extraction and affecting the integrity and accuracy of the knowledge base; on the other hand, the existing knowledge base update mechanism is often not flexible enough to timely reflect the dynamic changes of Internet of Things data, making the knowledge base lack timeliness and difficult to meet the actual application requirements.

[0003] Chinese Patent No. CN119357444A discloses a method, device, equipment and storage medium for updating a question-answering knowledge base, which cannot process according to the characteristics of Internet of Things big data and cannot effectively handle the massive, real-time and diverse nature of Internet of Things data. At the same time, it lacks an effective strategy for constructing and optimizing the knowledge base using historical data, and does not fully consider the optimization effect of knowledge base usage records on knowledge generation strategies, making it difficult to achieve continuous optimization and real-time update of the knowledge base.

[0004] In summary, there is an urgent need for a new technical solution for knowledge base processing based on Internet of Things big data to solve the above problems. Summary of the Invention

[0005] The purpose of the present application is to provide a knowledge base processing method and system based on Internet of Things big data to solve the technical problems proposed in the above background art.

[0006] To achieve the above purpose, the present application discloses the following technical solutions:

[0007] In the first aspect, the present application discloses a knowledge base processing method based on Internet of Things big data, and the method includes the following steps:

[0008] S1: Collect historical Internet of Things big data, preprocess the historical Internet of Things big data and extract key data features, use a preset initial knowledge generation strategy to perform knowledge mining on the key data features, generate knowledge entries, and construct an initial knowledge base; wherein, the preprocessing at least includes cleaning and denoising for removing error data, duplicate data, and outliers, the initial knowledge generation strategy is obtained based on initial knowledge requirements, and the initial knowledge generation strategy at least includes an initial knowledge mining direction and initial algorithm parameters, and the knowledge mining at least includes association analysis and trend prediction;

[0009] S2: Collect the usage records of the user for the initial knowledge base in real time, and the usage records are used to at least include information such as user questions, obtained knowledge content, usage time, and usage frequency;

[0010] S3: Extract the features of the usage records to generate usage record features, and adjust the initial knowledge generation strategy based on the usage record features to generate a real-time knowledge generation strategy; wherein, the real-time knowledge generation strategy at least includes an adjusted real-time knowledge mining direction and real-time algorithm parameters;

[0011] S4: Use the real-time knowledge generation strategy to perform real-time processing on the collected real-time Internet of Things big data to generate real-time knowledge; wherein, the real-time processing at least includes real-time association analysis and real-time trend prediction;

[0012] S5: Update the initial knowledge base with the real-time knowledge, and the update at least includes: for the newly generated knowledge entries, judge their relevance and complementarity with the existing knowledge in the knowledge base, and directly add them to the knowledge base when they are brand-new knowledge, and integrate the relevant knowledge when they are related to the existing knowledge.

[0013] Preferably, in S2, when collecting the usage records, collect direct knowledge query behavior records and knowledge usage situations involved in each link of the business process, including:

[0014] Collect n links in the business process, and in the i-th link of the n links, collect the number of times C of knowledge usage i and the number of knowledge categories T used i , then the usage record feature of this business process link is:

[0015]

[0016] wherein, B i is the usage record feature of the i-th link calculated.

[0017] Preferably, when analyzing the usage records, calculate the association features of knowledge usage in different application scenarios, including:

[0018] Obtain the existing m application scenarios. For scenario j and scenario k, count the number of knowledge N used simultaneously in scenario j and scenario k jk , the total number of knowledge used in scenario j is N j , the total number of knowledge used in scenario k is N k , then the knowledge usage association feature between scenario j and scenario k is:

[0019]

[0020] where, R jk is the calculated knowledge usage association feature between scenario j and scenario k, and analyze to obtain the knowledge usage association feature threshold R jk corresponding to R jk_τ , when R jk ≥R jk_τ , it is determined that the knowledge usage association degree between scenario j and scenario k is high. Then, when optimizing the real-time Internet of Things big data collection, when specific type of data is collected in one scenario, increase the amount of relevant data collected in the other scenario.

[0021] Preferably, in S3, detect the knowledge query behavior sequences that frequently appear in the usage records, including:

[0022] There are p behaviors in the knowledge query behavior sequence, and the occurrence time of each behavior is t1, t2,..., t p , calculate the time interval feature I of this sequence i =t i+1 -t i (i = 1, 2,..., i - 1), count the time interval features of all the same behavior sequences, and calculate the corresponding average time interval:

[0023]

[0024] where, q is the number of times the same behavior sequence appears, α is the average time interval correction parameter obtained based on regression analysis, and this average time interval correction parameter is used to represent the time required for the user to digest knowledge, is the calculated average time interval, and analyze to obtain the corresponding average time interval threshold When , it is determined that the user has urgency in obtaining knowledge of the knowledge type corresponding to this knowledge query behavior. When collecting Internet of Things big data in real time, increase the data collection frequency related to the knowledge of this knowledge type.

[0025] Preferably, when processing the real-time collected data, in combination with the average time interval, the data related to the knowledge of the knowledge type with urgency is preferentially processed, including:

[0026] The processing efficiency of the initial data cleaning algorithm for the data is E. For the data to be preferentially processed, the algorithm parameters are adjusted so that the processing efficiency E satisfies the constraint condition:

[0027]

[0028] where E0 is the processing efficiency of ordinary data, is the minimum value among all average time intervals, and the algorithm parameters that meet the constraint conditions are output and used to process the data that needs to be preferentially processed.

[0029] Preferably, in S4, when collecting real-time Internet of Things big data, based on the usage frequencies of different knowledge in the usage records, weight distribution is performed on the collected data, including:

[0030] If the usage frequency of the collected knowledge in the usage record is f, then the weight of the data collected related to this type of knowledge where f i is the usage frequency of all knowledge categories, and s is the total number of knowledge categories.

[0031] Preferably, when updating the knowledge base, it is determined whether to integrate based on the generated data weight and knowledge association degree corresponding to the newly generated knowledge entry, including:

[0032] When the generated data weight is greater than the preset generated data weight threshold and the knowledge association degree with the existing knowledge in the initial knowledge base is greater than the preset knowledge association degree threshold, this knowledge entry is integrated;

[0033] When integrating, the knowledge content and the association relationship between knowledge are correspondingly updated, and the update of this association relationship includes at least an increase in the association relationship.

[0034] Preferably, in S3, when analyzing the usage records, calculate the knowledge usage preference characteristics of different users and correspondingly adjust the collection amount, including:

[0035] For the v-th user among u users, count the proportion P of various types of knowledge used by this user vj , where j is the knowledge category, and based on this proportion, construct the knowledge usage preference feature vector of this user. When the difference in knowledge usage preferences between different users is greater than or equal to the preset difference threshold, when collecting real-time Internet of Things big data, data related to their preferences is collected separately for different users, and the calculation of its corresponding collection amount is:

[0036]

[0037] Among them, D0 is the basic acquisition volume, and D j is the acquisition volume of the data corresponding to the knowledge category j obtained by calculation.

[0038] Preferably, when generating the real-time knowledge, different knowledge generation strategies are adopted for different knowledge usage preference features, including:

[0039] Adjust the complexity of the data analysis algorithm by using the analysis depth adjustment formula, and adopt the corresponding knowledge generation strategy based on this complexity; among them, the analysis depth adjustment formula is:

[0040] d = d0 * (1 + Σ j∈Ω P vj )

[0041] wherein, d0 is the ordinary analysis depth, Ω is the set of knowledge categories preferred by this user, and d is the complexity of the data analysis algorithm obtained by calculation.

[0042] In a second aspect, the present application discloses a knowledge base processing system based on Internet of Things big data. This system is applicable to the knowledge base processing method based on Internet of Things big data as described above. This system includes:

[0043] An initial knowledge base construction module, and the initial knowledge base construction module is configured to: collect historical Internet of Things big data, preprocess the historical Internet of Things big data and extract key data features, and use a preset initial knowledge generation strategy to perform knowledge mining on the key data features, generate knowledge entries, and construct an initial knowledge base; wherein, the preprocessing at least includes cleaning and denoising for removing error data, duplicate data, and outliers, the initial knowledge generation strategy is obtained based on the initial knowledge requirement, and this initial knowledge generation strategy at least includes an initial knowledge mining direction and initial algorithm parameters, and the knowledge mining at least includes association analysis and trend prediction;

[0044] A usage record acquisition module, and the usage record acquisition module is configured to: collect in real time the usage records of the user for the initial knowledge base, and the usage records are used for information including at least user questions, obtained knowledge content, usage time, and usage frequency;

[0045] A real-time knowledge generation strategy generation module, and the real-time knowledge generation strategy generation module is configured to: extract the features of the usage records to generate usage record features, and adjust the initial knowledge generation strategy based on the usage record features to generate a real-time knowledge generation strategy; wherein, the real-time knowledge generation strategy at least includes an adjusted real-time knowledge mining direction and real-time algorithm parameters;

[0046] A real-time knowledge generation module, configured to: use the real-time knowledge generation strategy to perform real-time processing on the collected real-time Internet of Things big data to generate real-time knowledge; wherein, the real-time processing at least includes real-time association analysis and real-time trend prediction;

[0047] A knowledge base update module, configured to: use the real-time knowledge to update the initial knowledge base, and the update at least includes: for the newly generated knowledge entries, judge their relevance and complementarity with the existing knowledge in the knowledge base, and directly add them to the knowledge base when they are brand-new knowledge, and integrate the relevant knowledge when they are related to the existing knowledge.

[0048] Beneficial effects: The knowledge base processing method and system based on Internet of Things big data of the present application realize the effective integration and knowledge transformation of historical data, build a solid foundation for the knowledge base, collect the usage records of the initial knowledge base by users in real time, analyze their characteristics and adjust the initial knowledge generation strategy to obtain the real-time knowledge generation strategy, so that the knowledge generation can meet the actual needs of users, use the real-time knowledge generation strategy to process the real-time Internet of Things big data to generate real-time knowledge and update the initial knowledge base, achieve the dynamic update of the knowledge base, ensure the timeliness and accuracy of the knowledge base, effectively solve the problems of incomplete construction, untimely update of the knowledge base and difficulty in adapting to the characteristics of Internet of Things big data, and provide more accurate and effective knowledge support for Internet of Things applications. Brief Description of the Drawings

[0049] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative efforts.

[0050] Figure 1 It is a flowchart of the knowledge base processing method based on Internet of Things big data provided by the embodiment of the present application;

[0051] Figure 2 It is a structural block diagram of the knowledge base processing system based on Internet of Things big data provided by the embodiment of the present application. Detailed Embodiments

[0052] The following will clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, rather than all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0053] In this text, the term "including" is intended to cover non-exclusive inclusion, such that a process, method, article, or device that includes a series of elements not only includes those elements but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article, or device. Without further limitation, an element defined by the statement "including..." does not exclude the presence of additional identical elements in the process, method, article, or device that includes the said element.

[0054] The first aspect of this embodiment discloses a knowledge base processing method based on Internet of Things (IoT) big data as shown in Figure 1 which comprises the following steps:

[0055] S1: Collect historical IoT big data, preprocess the historical IoT big data and extract key data features, and use a preset initial knowledge generation strategy to perform knowledge mining on the key data features to generate knowledge entries and construct an initial knowledge base; wherein, the preprocessing at least includes cleaning and denoising for removing error data, duplicate data, and outliers, the initial knowledge generation strategy is obtained based on initial knowledge requirements, and the initial knowledge generation strategy at least includes an initial knowledge mining direction and initial algorithm parameters, and the knowledge mining at least includes association analysis and trend prediction;

[0056] S2: Collect in real time the usage records of the initial knowledge base, and the usage records are used for information that at least includes user questions, acquired knowledge content, usage time, and usage frequency;

[0057] S3: Extract the features of the usage records to generate usage record features, and adjust the initial knowledge generation strategy based on the usage record features to generate a real-time knowledge generation strategy; wherein, the real-time knowledge generation strategy at least includes an adjusted real-time knowledge mining direction and real-time algorithm parameters;

[0058] S4: Use the real-time knowledge generation strategy to perform real-time processing on the collected real-time IoT big data to generate real-time knowledge; wherein, the real-time processing at least includes real-time association analysis and real-time trend prediction;

[0059] S5: Use the real-time knowledge to update the initial knowledge base, and the update at least includes: for the newly generated knowledge entries, judge their relevance and complementarity with the existing knowledge in the knowledge base, and directly add them to the knowledge base when they are brand-new knowledge, and integrate the relevant knowledge when they are related to the existing knowledge.

[0060] It should be noted that this embodiment utilizes existing big data processing technologies to achieve the collection and preprocessing of IoT big data and the corresponding usage records.

[0061] Through the above, this embodiment realizes the effective integration of historical data and knowledge transformation, laying a solid foundation for the knowledge base. By collecting the usage records of the initial knowledge base by users in real time, analyzing their characteristics, and adjusting the initial knowledge generation strategy to obtain a real-time knowledge generation strategy, the knowledge generation can meet the actual needs of users. Using the real-time knowledge generation strategy to process real-time Internet of Things big data to generate real-time knowledge and update the initial knowledge base, the dynamic update of the knowledge base is achieved, ensuring the timeliness and accuracy of the knowledge base, effectively solving the problems of incomplete construction, untimely update, and difficulty in adapting to the characteristics of Internet of Things big data of the knowledge base, and providing more accurate and effective knowledge support for Internet of Things applications.

[0062] Specifically, in S2, when collecting usage records, directly collect the knowledge query behavior records and the knowledge usage situations involved in each link of the business process, including:

[0063] Collect n links in the business process. In the i-th link of the n links, collect the number of times C of knowledge usage i , the number of knowledge categories T used i , then the usage record feature of this business process link is:

[0064]

[0065] Among them, B i is the usage record feature of the i-th link calculated.

[0066] Through the above, this embodiment calculates the usage record features by using the number of times of knowledge usage and the number of knowledge categories in each link of the business process, realizing a detailed quantitative analysis of the knowledge usage situation in the business process. It can be understood that different from the existing method that only focuses on direct query behavior, this embodiment's design based on this reflects more comprehensively the user's needs and application of knowledge in actual business operations. The usage record features obtained in this way are more representative, can provide more accurate data support for subsequent adjustment of the knowledge generation strategy, help optimize the collection direction of real-time Internet of Things big data, make the collected data more in line with the actual business needs, thereby improving the matching degree between the knowledge in the knowledge base and the business scenario, and enhancing the practicality of the knowledge base in actual business.

[0067] Specifically, when analyzing usage records, calculate the correlation features of knowledge usage in different application scenarios, including:

[0068] Obtain m existing application scenarios. For scenarios j and k, count the number of knowledge N used simultaneously in scenarios j and k jk , the total number of knowledge used in scenario j is N j , the total number of knowledge used in scenario k is N k, the knowledge usage association feature between scenario j and scenario k is as follows:

[0069]

[0070] where, R jk is the calculated knowledge usage association feature between scenario j and scenario k, and by analyzing, the knowledge usage association feature threshold R jk corresponding to it is obtained. When R jk_τ ≥R jk jk_τ , it is determined that the knowledge usage association degree between scenario j and scenario k is high. Then, when optimizing the real-time IoT big data collection, when specific type of data is collected in one scenario, the collection volume of relevant data in another scenario is increased.

[0071] Through the above, this embodiment realizes the in-depth mining of the knowledge association relationship between different scenarios by using the knowledge usage association feature under different application scenarios. In a specific application, the knowledge usage association feature threshold is obtained by using the existing regression analysis, and based on this threshold, the level of the knowledge usage association degree between scenarios is accurately judged. In an actual example, when the association degree is high, the collection volume of relevant scenario data is increased when collecting real-time IoT big data, avoiding the blindness of data collection, improving the pertinence and effectiveness of data collection, so that the knowledge base can cover more comprehensive and interrelated knowledge, better meet the knowledge needs of users in different application scenarios, and enhance the comprehensive service ability and application value of the knowledge base.

[0072] Specifically, in S3, the knowledge query behavior sequences frequently appearing in the usage records are detected, including:

[0073] There are p behaviors in the knowledge query behavior sequence, and the occurrence times of each behavior are t1, t2,..., t p in sequence. Calculate the time interval feature I i =t i+1 -t i (i = 1, 2,..., i - 1), and count the time interval features of all the same behavior sequences, and calculate the corresponding average time interval:

[0074]

[0075] where, q is the number of times the same behavior sequence appears, and α is the average time interval correction parameter obtained by regression analysis. This average time interval correction parameter is used to represent the time required for users to digest knowledge. is the calculated average time interval, and by analyzing, the average time interval threshold corresponding to it is obtained. When When ​When it is determined that the user has an urgent need to acquire knowledge of the knowledge type corresponding to the knowledge query behavior, the data collection frequency related to the knowledge of this knowledge type is increased when collecting IoT big data in real time.

[0076] It can be understood that the ultimate goal of building a knowledge base based on IoT big data is to provide more accurate knowledge for users' decision-making. Based on this goal, combined with the average time interval designed in this embodiment to evaluate the urgency of knowledge acquisition of the knowledge type corresponding to the user's knowledge query behavior and the actual application, this embodiment designs an average time interval correction parameter to correct the average time interval, so as to make the analysis of urgency more accurate and avoid misjudgment of urgency caused by repeated operations.

[0077] Through the above, this embodiment detects the knowledge query behavior sequence that frequently appears in the usage records, calculates its time interval characteristics and average time interval, and uses the existing regression analysis to set the average time interval threshold to judge the user's urgency for knowledge acquisition, achieving an accurate grasp of the timeliness of the user's knowledge needs. According to the judgment result, the data collection frequency related to the urgently needed knowledge is increased, and the latest data can be obtained in time for knowledge generation, thus ensuring that the knowledge base can quickly respond to the user's urgent knowledge needs, provide more timely and effective knowledge services, enhance the application ability of the knowledge base in scenarios with high real-time requirements, and avoid affecting the user's decision-making due to untimely knowledge update.

[0078] Specifically, when processing the real-time collected data, combined with the average time interval, the data related to the knowledge of the knowledge type with urgency is preferentially processed, including:

[0079] The processing efficiency of the initial data cleaning algorithm for data is E. For the data to be preferentially processed, the algorithm parameters are adjusted so that the processing efficiency E satisfies the constraint condition:

[0080]

[0081] where E0 is the processing efficiency of ordinary data, is the minimum value among all average time intervals, and the algorithm parameters that meet the constraint conditions are output and used to process the data that needs to be preferentially processed.

[0082] With the above, this embodiment gives priority to processing data related to knowledge types with urgency by combining the average time interval, and improves the processing efficiency by adjusting algorithm parameters, achieving efficient utilization of real-time collected data. In practical applications, giving priority to processing urgent demand data can quickly extract valuable information from the data to generate real-time knowledge, meeting the user's requirements for the timeliness of knowledge. Further, improving the processing efficiency can reduce the time cost of data processing, accelerate the update speed of the knowledge base, enable the knowledge base to provide the latest and most accurate knowledge to users in a shorter time, and improve the performance of the knowledge base in terms of timeliness and efficiency. And only giving priority to processing urgent demand data realizes the reasonable allocation of computing power resources first.

[0083] Specifically, in S4, when collecting real-time Internet of Things big data, based on the usage frequencies of different knowledge in the usage records, weight assignment is performed on the collected data, including:

[0084] If the usage frequency of the collected knowledge in the usage record is f, the weight of the data related to this type of knowledge collected where f i is the usage frequency of all knowledge categories, and s is the total number of knowledge categories.

[0085] With the above, this embodiment realizes a reasonable evaluation of the importance of data by performing weight assignment on the collected data based on the usage frequencies of different knowledge in the usage records. In practical applications, different processing methods are adopted for different data based on the weights. More accurate data mining algorithms are used for data with high weights, which can more accurately generate real-time knowledge from high-value data, so that the knowledge base can focus more on the knowledge fields frequently used by users, improve the quality and pertinence of knowledge generation, avoid wasting too many resources on low-value data, optimize the resource utilization efficiency of the knowledge base, and improve the practicality and reliability of the knowledge base.

[0086] Specifically, when updating the knowledge base, it is judged whether to integrate based on the generated data weight and knowledge correlation degree corresponding to the newly generated knowledge entry, including:

[0087] When the generated data weight is greater than the preset generated data weight threshold and the knowledge correlation degree with the existing knowledge in the initial knowledge base is greater than the preset knowledge correlation degree threshold, integrate this knowledge entry;

[0088] When integrating, update the knowledge content and the association relationship between knowledge correspondingly. The update of this association relationship includes at least the increase of the association relationship.

[0089] With the above, this embodiment determines whether to integrate based on the generated data weight and knowledge association degree corresponding to the newly generated knowledge entry, and updates the knowledge content and association relationship when the conditions are met, realizing the intelligent update of the knowledge base and the optimization of the knowledge structure. By setting the corresponding generated data weight threshold and knowledge association degree threshold based on the common knowledge known to those skilled in the art, valuable knowledge can be screened out for integration, avoiding redundancy and confusion of knowledge in the knowledge base. Updating the knowledge association relationship can make the knowledge in the knowledge base form a closer network, facilitating users' knowledge retrieval and associated learning, improving the knowledge organization and management level of the knowledge base, and enhancing the knowledge service ability of the knowledge base. In a simple example, the number of association relationships of the knowledge before update is N1, and the number of association relationships of the newly generated knowledge entry is N2, then the number of association relationships of the updated and integrated knowledge is N1 + N2.

[0090] Specifically, in S3, when analyzing the usage records, calculate the knowledge usage preference characteristics of different users and adjust the collection amount accordingly, including:

[0091] For the v-th user among u users, count the proportion P of various types of knowledge used by this user vj , where j is the knowledge category, and based on this proportion, construct the knowledge usage preference feature vector of this user. When the difference in knowledge usage preferences between different users is greater than or equal to the preset difference threshold, when collecting real-time Internet of Things big data, collect data related to their preferences for different users respectively, and the calculation of the corresponding collection amount is:

[0092]

[0093] where D0 is the basic collection amount, and D j is the collection amount of the data corresponding to the knowledge category j calculated.

[0094] With the above, this embodiment realizes the precise satisfaction of the personalized needs of different users by calculating the knowledge usage preference characteristics of different users and adjusting the collection amount according to the preference difference. Collecting data related to their preferences for different users respectively can ensure that the collected data better meets the actual needs of users, so that the knowledge base can provide customized knowledge services for different users, improving users' satisfaction and dependence on the knowledge base. Under the guidance of the difference threshold formulated based on the common knowledge known to those skilled in the art, adjusting the collection amount based on preferences can avoid over-collection and waste of data, optimize the resource allocation of data collection, and enhance the ability of the knowledge base in personalized services.

[0095] Specifically, when generating real-time knowledge, different knowledge generation strategies are adopted according to different knowledge usage preference characteristics, including:

[0096] Adjust the complexity of the data analysis algorithm using the analysis depth adjustment formula, and adopt the corresponding knowledge generation strategy based on this complexity; among them, the analysis depth adjustment formula is:

[0097] d = d0 * (1 + Σ j∈Ω P vj )

[0098] Among them, d0 is the ordinary analysis depth, Ω is the set of knowledge categories preferred by this user, and d is the calculated complexity of the data analysis algorithm.

[0099] Through the above, this embodiment uses different knowledge generation strategies according to different knowledge usage preference characteristics, and adjusts the complexity of the data analysis algorithm through the analysis depth adjustment formula, realizing the personalization and precision of knowledge generation. In this embodiment, the premise of adopting different analysis depths and knowledge generation strategies for data with different user preferences is the analysis of complexity. Based on this analysis, knowledge that better meets the user's needs is mined from the data, so that the knowledge base can provide more professional and in-depth knowledge services for users with different preferences, improving the ability of the knowledge base to meet the diverse needs of users, and further enhancing the competitiveness and application value of the knowledge base.

[0100] The second aspect of this embodiment discloses a Figure 2 knowledge base processing system based on Internet of Things big data as shown. This system is applicable to the above knowledge base processing method based on Internet of Things big data. This system includes:

[0101] Initial knowledge base construction module. The initial knowledge base construction module is configured to: collect historical Internet of Things big data, preprocess the historical Internet of Things big data and extract key data features, use the preset initial knowledge generation strategy to perform knowledge mining on the key data features, generate knowledge entries, and construct an initial knowledge base; among them, the preprocessing at least includes cleaning and denoising for removing error data, duplicate data, and outliers. The initial knowledge generation strategy is obtained based on the initial knowledge requirements, and this initial knowledge generation strategy at least includes the initial knowledge mining direction and initial algorithm parameters. The knowledge mining at least includes association analysis and trend prediction;

[0102] Usage record collection module. The usage record collection module is configured to: collect the usage records of users on the initial knowledge base in real time. The usage records are used for information at least including user questions, obtained knowledge content, usage time, and usage frequency;

[0103] A real-time knowledge generation strategy generation module, configured to: extract features of usage records to generate usage record features, adjust an initial knowledge generation strategy based on the usage record features, and generate a real-time knowledge generation strategy; wherein the real-time knowledge generation strategy at least includes an adjusted real-time knowledge mining direction and real-time algorithm parameters;

[0104] A real-time knowledge generation module, configured to: use the real-time knowledge generation strategy to perform real-time processing on the collected real-time Internet of Things big data to generate real-time knowledge; wherein the real-time processing at least includes real-time association analysis and real-time trend prediction;

[0105] A knowledge base update module, configured to: use the real-time knowledge to update the initial knowledge base, and the update at least includes: for newly generated knowledge entries, judge their relevance and complementarity with the existing knowledge in the knowledge base, directly add them to the knowledge base when they are brand-new knowledge, and integrate the relevant knowledge when they are related to the existing knowledge.

[0106] It should be noted that the knowledge base processing system based on Internet of Things big data in this embodiment corresponds to the foregoing knowledge base processing method based on Internet of Things big data. Therefore, the content not specifically described in the knowledge base processing system based on Internet of Things big data in this embodiment, such as but not limited to function definitions, working principles, and technical effects, etc., can refer to the descriptions in the foregoing knowledge base processing method based on Internet of Things big data, and will not be elaborated herein.

[0107] In summary, the knowledge base processing method and system based on Internet of Things big data in this embodiment realize the effective integration and knowledge transformation of historical data, build a solid foundation for the knowledge base. By collecting the usage records of the initial knowledge base by users in real time, analyzing their features and adjusting the initial knowledge generation strategy to obtain the real-time knowledge generation strategy, the knowledge generation can meet the actual needs of users. Using the real-time knowledge generation strategy to process the real-time Internet of Things big data to generate real-time knowledge and update the initial knowledge base, the dynamic update of the knowledge base is achieved, ensuring the timeliness and accuracy of the knowledge base, effectively solving the problems of incomplete construction, untimely update, and difficulty in adapting to the characteristics of Internet of Things big data of the knowledge base, and providing more accurate and effective knowledge support for Internet of Things applications.

[0108] In the embodiments provided in the present application, it should be understood that the embodiments described herein can be implemented in hardware, software, firmware, middleware, code, or any appropriate combination thereof. For hardware implementation, the processor can be implemented in one or more of the following units: application specific integrated circuit (ASIC), digital signal processor (DSP), digital signal processing device (DSPD), programmable logic device (PLD), field programmable gate array (FPGA), processor, controller, microcontroller, microprocessor, other electronic units designed to implement the functions described herein, or a combination thereof. For software implementation, part or all of the processes of the embodiments can be completed by instructing the relevant hardware through a computer program. When implemented, the above program can be stored in a computer-readable storage medium or transmitted as one or more instructions or codes on a computer-readable storage medium. The computer-readable storage medium includes computer storage media and communication media, where the communication media includes any medium that facilitates the transfer of a computer program from one place to another. The storage media can be any available medium that can be accessed by a computer. The computer-readable storage medium can include, but is not limited to, RAM, ROM, EEPROM, CD-ROM, or other optical disk storage, magnetic disk storage media, or other magnetic storage devices, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer.

[0109] Finally, it should be noted that the above are only the preferred embodiments of the present application and are not used to limit the present application. Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A knowledge base processing method based on Internet of Things big data, characterized in that The method includes the following steps: S1: Collect historical Internet of Things big data, preprocess the historical Internet of Things big data and extract key data features, use a preset initial knowledge generation strategy to perform knowledge mining on the key data features, generate knowledge entries, and construct an initial knowledge base; wherein, the preprocessing at least includes cleaning and denoising for removing error data, duplicate data, and outliers, the initial knowledge generation strategy is obtained based on initial knowledge requirements, and the initial knowledge generation strategy at least includes an initial knowledge mining direction and initial algorithm parameters, and the knowledge mining at least includes association analysis and trend prediction; S2: Real-time collect the usage records of the initial knowledge base by users, and the usage records are used to at least include information such as user questions, obtained knowledge content, usage time, and usage frequency; S3: Extract the features of the usage records to generate usage record features, and adjust the initial knowledge generation strategy based on the usage record features to generate a real-time knowledge generation strategy; wherein, the real-time knowledge generation strategy at least includes an adjusted real-time knowledge mining direction and real-time algorithm parameters; S4: Use the real-time knowledge generation strategy to perform real-time processing on the collected real-time Internet of Things big data to generate real-time knowledge; wherein, the real-time processing at least includes real-time association analysis and real-time trend prediction; S5: Use the real-time knowledge to update the initial knowledge base, and the update at least includes: for newly generated knowledge entries, judge their relevance and complementarity with the existing knowledge in the knowledge base, directly add them to the knowledge base when they are brand-new knowledge, and integrate the relevant knowledge when they are related to the existing knowledge.

2. The knowledge base processing method based on Internet of Things big data according to claim 1, wherein, In S2, when collecting the usage records, collect direct knowledge query behavior records and knowledge usage situations involved in each link of the business process, including: n links in the acquisition business process. In the i-th link among the n links, the number of times C that knowledge is used in acquisition i and the number of knowledge categories T used i . Then the usage record feature of this business process link is: Among them, B i is the usage record feature of the i-th link calculated.

3. The knowledge base processing method based on Internet of Things big data according to claim 2, wherein When analyzing the usage records, calculate the association features of knowledge usage in different application scenarios, including: Obtain the existing m application scenarios. For scenario j and scenario k, count the number of knowledge N that is used in both scenario j and scenario k jk , the total number of knowledge used in scenario j is N j , the total number of knowledge used in scenario k is N k , then the knowledge usage association feature between scenario j and scenario k is: Among them, R jk is the knowledge usage correlation feature between the calculated scenario j and scenario k, and the obtained R jk corresponds to the knowledge usage correlation feature threshold R jk_τ . When R jk ≥R jk_τ , it is determined that the knowledge usage correlation degree between scenario j and scenario k is high. Then, when optimizing the real-time Internet of Things big data collection, when a specific type of data is collected in one scenario, the collection volume of related data in another scenario is increased.

4. The knowledge base processing method based on Internet of Things big data according to claim 1, characterized in that In S3, detect the frequently occurring knowledge query behavior sequences in the usage records, including: There are p behaviors in the knowledge query behavior sequence, and the occurrence times of each behavior are t1, t2,..., t p , calculate the time interval feature I of this sequence i = t i+1 - t i (i = 1, 2,..., i - 1), count the time interval features of all the same behavior sequences, and calculate the corresponding average time interval: Where q is the number of occurrences of the same behavior sequence, and α is the average time interval correction parameter obtained based on regression analysis. This average time interval correction parameter is used to represent the time required for users to digest knowledge. is the calculated average time interval, and is obtained through analysis the corresponding average time interval threshold When it is determined that the user has an urgency to acquire knowledge of the knowledge type corresponding to the knowledge query behavior. When collecting IoT big data in real time, the data collection frequency related to the knowledge of this knowledge type is increased.

5. The knowledge base processing method based on Internet of Things big data according to claim 4, wherein When processing the real-time collected data, in combination with the average time interval, preferentially process the relevant data of the knowledge of the knowledge types with urgency, including: The processing efficiency of the initial data cleaning algorithm for data is E. For the data to be preferentially processed, adjust the algorithm parameters so that the processing efficiency E meets the constraint conditions: Among them, E0 is the processing efficiency of ordinary data, which is the minimum value among all average time intervals, outputs algorithm parameters that meet the constraint conditions and is used to process the data that needs to be processed preferentially.

6. The knowledge base processing method based on Internet of Things big data according to claim 1, characterized in that In S4, when collecting real-time Internet of Things big data, based on the usage frequencies of different knowledge in the usage records, perform weight assignment on the collected data, including: If the usage frequency of the collected knowledge in the usage record is f, the data weight related to this type of knowledge collected where f i is the usage frequency of all knowledge categories, and s is the total number of knowledge categories.

7. The knowledge base processing method based on Internet of Things big data according to claim 6, characterized in that, When updating the knowledge base, judge whether to integrate based on the generated data weight and knowledge correlation degree corresponding to the newly generated knowledge entries, including: Integrate the knowledge entry when the generated data weight is greater than a preset generated data weight threshold and the knowledge correlation degree with the existing knowledge in the initial knowledge base is greater than a preset knowledge correlation degree threshold; When integrating, update the knowledge content and the association relationship between knowledge correspondingly, and the update of the association relationship at least includes the addition of the association relationship.

8. The knowledge base processing method based on Internet of Things big data according to claim 1, characterized in that In S3, when analyzing the usage records, calculate the knowledge usage preference features of different users and correspondingly adjust the collection amount, including: For the v-th user among u users, count the proportion P of various types of knowledge used by this user vj , where j is the knowledge category, and based on this proportion, construct the knowledge usage preference feature vector of this user. When the difference in knowledge usage preferences between different users is greater than or equal to a preset difference threshold, collect data related to their preferences separately for different users when collecting real-time IoT big data, and the calculation of the corresponding collection volume is as follows: Among them, D0 is the basic acquisition volume, and D j is the acquisition volume of the data corresponding to the knowledge category j obtained by calculation.

9. The knowledge base processing method based on Internet of Things big data according to claim 8, characterized in that, When generating the real-time knowledge, different knowledge generation strategies are adopted according to different knowledge usage preference features, including: Adjusting the complexity of the data analysis algorithm by using the analysis depth adjustment formula, and adopting the corresponding knowledge generation strategy based on this complexity; where the analysis depth adjustment formula is: d = d0 * (1 + Σ j∈Ω P vj ) where d0 is the ordinary analysis depth, Ω is the set of knowledge categories preferred by this user, and d is the calculated complexity of the data analysis algorithm.

10. A knowledge base processing system based on Internet of Things big data, which is applicable to the knowledge base processing method based on Internet of Things big data described in any one of claims 1-9, and is characterized in that, The system includes: An initial knowledge base construction module, configured to: collect historical Internet of Things big data, preprocess the historical Internet of Things big data and extract key data features, use a preset initial knowledge generation strategy to perform knowledge mining on the key data features, generate knowledge entries, and construct an initial knowledge base; where the preprocessing at least includes cleaning and denoising for removing error data, duplicate data, and outliers, the initial knowledge generation strategy is obtained based on the initial knowledge requirement, and the initial knowledge generation strategy at least includes the initial knowledge mining direction and initial algorithm parameters, and the knowledge mining at least includes association analysis and trend prediction; A usage record collection module, configured to: collect in real time the usage records of the user for the initial knowledge base, and the usage records are used to include at least information such as user questions, obtained knowledge content, usage time, and usage frequency; A real-time knowledge generation strategy generation module, configured to: extract the features of the usage records to generate usage record features, and adjust the initial knowledge generation strategy based on the usage record features to generate a real-time knowledge generation strategy; where the real-time knowledge generation strategy at least includes the adjusted real-time knowledge mining direction and real-time algorithm parameters; A real-time knowledge generation module, configured to: use the real-time knowledge generation strategy to perform real-time processing on the collected real-time Internet of Things big data to generate real-time knowledge; where the real-time processing at least includes real-time association analysis and real-time trend prediction; A knowledge base update module, configured to: update the initial knowledge base with the real-time knowledge, and the update at least includes: for the newly generated knowledge entries, judging their relevance and complementarity with the existing knowledge in the knowledge base, directly adding them to the knowledge base when they are new knowledge, and integrating the relevant knowledge when they are related to the existing knowledge.

Images