Patent data scheduling method and system

Abstract

The invention provides a patent data scheduling method and system, and belongs to the technical field of computers. Historical CPU utilization rate data and historical memory utilization rate data of each computing resource in a patent data scheduling system are used as basic data, an iteration model is input, a prediction result is output and compared with a preset threshold value, and a comparison result is used as a basis for patent data scheduling task priority adjustment and task pause. The problems that in the prior art, data scheduling efficiency is reduced and energy consumption is increased in a patent or intellectual property data scheduling system can be solved at least to a certain extent.

Description

Technical Field

[0001] The present invention belongs to the field of computer technology, and in particular, relates to a patented data scheduling method and system. Background Art

[0002] As a form of intellectual property, patents have become one of the infrastructures of the business world. Patent data scheduling plays a vital role in helping service recipients understand the current status of technological development, predict development trends, determine R&D directions, track competitor R&D directions, and monitor risks.

[0003] In the existing patent data scheduling method, the patent with publication number CN109189942B involves the construction of an intellectual property graph. The patent data is preprocessed and the patent data of the same type after merging are segmented to obtain the segmented data of each type of patent data; the RAKE algorithm is used to extract knowledge from the preprocessed patent data, and the segmented data of each type of patent data is cleaned to obtain the corresponding subject original file, and keywords are extracted to obtain subject words, and a patent subject library is constructed for each type of patent data; the entities of the patent data are defined, the subjects of the patent data are determined, the entities and subjects of the patent are identified according to the general knowledge graph, the semantic association relationship between the entities and the subjects is mined, and the patent data knowledge graph is constructed. The patent with publication number CN112561380A involves intellectual property scheduling. It uses a business allocation management module to view the real-time progress of the business assigned to different engineers and assign new tasks so that engineers can have time to handle related business; using the business copywriting module, engineers edit the content into core content and upload it to the local area network. The system performs preliminary automatic retrieval based on the core content, and matches the similarity between different retrieval methods and the core content of the business, thereby simplifying the basis for retrieval for engineers; using the database storage module, engineers can search in the database. After the system performs a preliminary search, engineers can continue a secondary search based on the preliminary search, thereby saving engineers' time.

[0004] The inventors have found through searching that, as shown in the above patents, most of the patent or intellectual property data scheduling methods in the prior art do not fully consider the CPU usage and memory usage of the patent or intellectual property data scheduling system, which can easily lead to a decrease in the data scheduling efficiency of the data scheduling system and an increase in energy consumption. Summary of the invention

[0005] In order to solve the problems existing in the prior art, the present invention proposes a patent data scheduling method to improve the data scheduling efficiency of the patent data scheduling system and reduce energy consumption, at least to a certain extent.

[0006] In order to achieve the above-mentioned purpose, the present invention adopts the following technical scheme: a patent data scheduling method, comprising the following steps: S1: monitoring the patent data scheduling task of each computing resource in the patent data scheduling system, and obtaining the historical CPU usage data and historical memory usage data of each computing resource; S2: preprocessing the historical CPU usage data and historical memory usage data to obtain a first intermediate data sequence D T =(x 1 , x 2 , x 3 , ..., x T ), T represents the total number of data in the first intermediate data sequence; S3: set the number of iterations, and convert the first intermediate data sequence D T =(x 1 , x 2 , x 3 , ..., x T ) is input into the iterative model for iteration to obtain the predicted data at time n, where n is less than or equal to T; S4: the predicted data at time n is compared with the preset data threshold, and the patent data scheduling task priority and scheduling task control of each computing resource are adjusted according to the comparison result to realize patent data scheduling.

[0007] Furthermore, in step S2, the historical CPU usage data and the historical memory usage data are preprocessed to obtain a first intermediate data sequence D T =(x 1 , x 2 , x 3 , ..., x T ), specifically comprising the following sub-steps: S21: performing differentiation and filtering processing on the above historical CPU usage data and historical memory usage data to obtain processed historical CPU usage data and processed historical memory usage data; S22: intercepting a first proportion of the processed historical CPU usage data, wherein the first proportion is preferably 80%; S23: intercepting a second proportion of the processed historical memory usage data, wherein the second proportion is preferably 70%; S24: combining the first proportion of the processed historical CPU usage data obtained in step S22 and the second proportion of the processed historical memory usage data obtained in step S23 as the first intermediate data sequence D T =(x 1 , x 2 , x 3 , ..., x T ).

[0008] Further, in the step S24, the processed historical CPU usage data of the first ratio obtained in step S22 and the processed historical memory usage data of the second ratio obtained in step S23 are combined, specifically, the processed historical CPU usage data of the first ratio obtained in step S22 is first, and the processed historical memory usage data of the second ratio obtained in step S23 is later, and the data are spliced ​​to obtain the first intermediate data sequence D T .

[0009] Furthermore, the iterative model in step S3 is specifically:

[0010]

[0011]

[0012]

[0013] s t-1 =k 1 ×random(0,1)

[0014] ITE t-1 =k 2 ×random(0,1);

[0015] Among them, LF t represents the logistic equation at time t, which is used as the threshold function of the iterative model. e represents a natural constant, avg() represents an average function, sinh() represents a hyperbolic sine function, cosh() represents a hyperbolic cosine function, random() represents a random number function, and x t represents the data at time t in the first intermediate data sequence, t = 2, 3, ..., n, s t and ITE t They represent the input data state variable and iteration state variable at time t, respectively, and s 1 and ITE 1 They represent the initial value of the input data state variable and the initial value of the iteration state variable respectively. The input data state variable at time t is used to clean the data after each iteration. The iteration state variable at time t is used to characterize the output data after each iteration. 1 and k 2 Represent the first adjustment coefficient and the second adjustment coefficient respectively.

[0016] Furthermore, in step S4, the predicted data at time n is compared with a preset data threshold, and the priority of the patent data scheduling task of each computing resource is adjusted and the scheduling task is controlled according to the comparison result, specifically: iteratively obtain the iteration state variable ITE n, and the iteration state variable ITE n The data at time n used as the prediction is compared with the preset data threshold, and the iterative state variable ITE is n When the data value is greater than or equal to the preset data threshold, the priority of the current patent data scheduling task is reduced and the current task execution is suspended; when the iteration state variable ITE n When it is less than the preset data threshold, the priority of the current patent data scheduling task remains unchanged.

[0017] The present invention also provides a patent data scheduling system, including a storage subsystem, a monitoring subsystem and a data fusion subsystem. The data fusion subsystem is used to integrate, clean and fuse patent data from different data sources, and send the processed patent data to the storage subsystem for storage of the patent data. The monitoring subsystem uses multi-threading technology to divide each computing resource by IP, and performs polling monitoring and control on the patent data scheduling tasks of each computing resource. The monitoring subsystem adjusts the priority of the patent data scheduling tasks and controls the scheduling tasks for each computing resource based on the above-mentioned patent data scheduling method to realize patent data scheduling.

[0018] The beneficial technical effect of the present invention compared with the prior art is that: based on the historical CPU usage data and historical memory usage data of each computing resource in the patent data scheduling system, an iterative model is input, a prediction result is output and compared with a preset threshold, and the comparison result is used as the basis for adjusting the priority of patent data scheduling tasks and pausing tasks. This can at least to a certain extent solve the problems of decreased data scheduling efficiency and increased energy consumption that are prone to occur in patent or intellectual property data scheduling systems in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the implementation methods of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the implementation methods or the description of the prior art. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived from the provided drawings without creative work.

[0020] Figure 1 The flowchart of the patent data scheduling method of the present invention;

[0021] Figure 2 This is a simplified structural diagram of the patented data scheduling system of the present invention. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] The following first describes the concepts involved in the present application in conjunction with the accompanying drawings. It should be noted that the following description of each concept is only to make the content of the present application easier to understand, and does not limit the scope of protection of the present application; at the same time, the embodiments and features in the embodiments of the present application can be combined with each other in the absence of conflict. The present application will be described in detail with reference to the accompanying drawings and in conjunction with the embodiments.

[0024] In conjunction with the instruction manual Figure 1 The present invention proposes a patent data scheduling method, which includes the following steps: S1: monitoring the patent data scheduling task of each computing resource in the patent data scheduling system, and obtaining the historical CPU usage data and historical memory usage data of each computing resource; S2: preprocessing the historical CPU usage data and historical memory usage data to obtain a first intermediate data sequence D T =(x 1 , x 2 , x 3 , ..., x T ), T represents the total number of data in the first intermediate data sequence; S3: set the number of iterations, and convert the first intermediate data sequence D T =(x 1 , x 2 , x 3 , ..., x T ) is input into the iterative model for iteration to obtain the predicted data at time n, where n is less than or equal to T; S4: the predicted data at time n is compared with the preset data threshold, and the patent data scheduling task priority and scheduling task control of each computing resource are adjusted according to the comparison result to realize patent data scheduling.

[0025] In step S2, the historical CPU usage data and the historical memory usage data are preprocessed to obtain a first intermediate data sequence D T =(x 1 , x 2 , x 3 , ..., x T), specifically including the following sub-steps: S21: performing differentiation and filtering processing on the above historical CPU usage data and historical memory usage data to obtain processed historical CPU usage data and processed historical memory usage data, and the present invention preferably adopts Kalman filtering; S22: intercepting a first proportion of the processed historical CPU usage data, and the first proportion is preferably 80%; S23: intercepting a second proportion of the processed historical memory usage data, and the second proportion is preferably 70%; S24: combining the first proportion of the processed historical CPU usage data obtained in step S22 and the second proportion of the processed historical memory usage data obtained in step S23 as the first intermediate data sequence D T =(x 1 , x 2 , x 3 , ..., x T ). Of course, the first ratio and the second ratio may be set to 80% or 70%, but should not be less than 70% to ensure that there is enough data input into the iterative model for training.

[0026] In the step S24, the processed historical CPU usage data of the first ratio obtained in step S22 and the processed historical memory usage data of the second ratio obtained in step S23 are combined, specifically, the processed historical CPU usage data of the first ratio obtained in step S22 is first, and the processed historical memory usage data of the second ratio obtained in step S23 is later, and the data are spliced ​​to obtain the first intermediate data sequence D T , as basic data input to the iterative model.

[0027] The iterative model in step S3 is specifically:

[0028]

[0029]

[0030]

[0031] s t-1 =k 1 ×random(0,1)

[0032] ITE t-1 =k 2 ×random(0,1);

[0033] Among them, LF trepresents the logistic equation at time t, which is used as the threshold function of the iterative model. e represents a natural constant, avg() represents an average function, sinh() represents a hyperbolic sine function, cosh() represents a hyperbolic cosine function, random() represents a random number function, and x t represents the data at time t in the first intermediate data sequence, t = 2, 3, ..., n, s t and ITE t They represent the input data state variable and iteration state variable at time t, respectively, and s 1 and ITE 1 They represent the initial value of the input data state variable and the initial value of the iteration state variable respectively. The input data state variable at time t is used to clean the data after each iteration. The iteration state variable at time t is used to characterize the output data after each iteration. 1 and k 2 The iterative model of the present invention cleans the data in each iteration, so that the iterative model can be effectively applied to the prediction of nonlinear data, improve the prediction accuracy, and further improve the scheduling efficiency of patent data.

[0034] In step S4, the predicted data at time n is compared with the preset data threshold, and the priority of the patent data scheduling task of each computing resource is adjusted and the scheduling task is controlled according to the comparison result. Specifically, the iteration state variable ITE is obtained according to the preset number of iterations. n , and the iteration state variable ITE n As the predicted data at time n, it is compared with the preset data threshold. When the iterative state variable ITE n When the data value is greater than or equal to the preset data threshold, the priority of the current patent data scheduling task is reduced and the current task execution is suspended to reduce the speed of data generation in the memory to avoid the accumulation of memory data. The alarm submodule can also be used to alarm the administrator. When the iteration state variable ITE n When it is less than the preset data threshold, the priority of the current patent data scheduling task is kept unchanged and the current task is continued. The preset number of iterations and the preset data threshold need to be determined according to the current total task of the patent data scheduling system.

[0035] In conjunction with the instruction manual Figure 2The present invention also provides a patent data scheduling system, including a storage subsystem, a monitoring subsystem and a data fusion subsystem. The data fusion subsystem is the core part of the entire system. Its main function is to integrate, clean and fuse data from different data sources, and provide unified and high-quality data support for subsequent patent data updates or additions. When designing the data fusion module, the present invention fully considers the differences in data formats and structures in different countries, and adopts a multi-level and multi-strategy data fusion method. First, data from different sources are pre-processed, including data cleaning, format conversion, etc., to ensure the accuracy and consistency of the data. Secondly, in order to ensure the final structure of the data, a unified intermediate data entity is adopted to solve the problems of data heterogeneity and data redundancy, and create conditions for the subsequent patent search to use unified fields, unified analysis, and unified echo.

[0036] After the data fusion subsystem integrates, cleans and fuses the patent data from different data sources, the processed patent data is sent to the storage subsystem for storage. The storage subsystem adopts the ElasticSearch distributed storage architecture to achieve distributed storage and efficient processing of massive patent data, which not only provides high availability and scalability, but also greatly improves the efficiency of data processing. In order to meet subsequent needs for patent retrieval, analysis, etc., the present invention also carefully designs the type or word segmentation strategy of the patent field.

[0037] The monitoring subsystem uses multi-threading technology to divide each computing resource by IP, and performs polling monitoring and control on the patent data scheduling task of each computing resource. The monitoring subsystem adjusts the priority of the patent data scheduling task and controls the scheduling task for each computing resource according to the patent data scheduling method proposed by the present invention to achieve patent data scheduling. The monitoring subsystem may also include a data display submodule and an alarm submodule. The data display submodule uses visualization technology to display the predicted data, which is convenient for intuitively understanding the changing trends and characteristics of the data. The alarm submodule is used to alarm the administrator of abnormal situations.

[0038] The patented data scheduling system and method of the present invention improve the data scheduling efficiency of the patented data scheduling system to at least a certain extent, reduce energy consumption, and have practical commercial application value.

[0039] The embodiments and / or implementation methods described above are only used to illustrate the preferred embodiments and / or implementation methods for realizing the technology of the present invention, and are not intended to limit the implementation methods of the technology of the present invention in any form. Any technical personnel in this field may make slight changes or modifications to other equivalent embodiments without departing from the scope of the technical means disclosed in the content of the present invention, but they should still be regarded as technologies or embodiments that are essentially the same as the present invention.

[0040] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and its core ideas of this application. The above is only the preferred implementation method of this application. It should be pointed out that due to the limitations of textual expression and the objective existence of infinite specific structures, ordinary technicians in this technical field can make several improvements, modifications or changes without departing from the principles of this application, and can also combine the above technical features in an appropriate manner; these improvements, modifications, changes or combinations, or the direct application of the concept and technical solution of the invention to other occasions without improvement, should be regarded as the scope of protection of this application.

Claims

1. A patented data scheduling method, characterized in that: The patented data scheduling method comprises the following steps: S1: monitor the patent data scheduling tasks of each computing resource in the patent data scheduling system, and obtain the historical CPU usage data and historical memory usage data of each computing resource; S2: Preprocess the historical CPU usage data and historical memory usage data to obtain the first intermediate data sequence D T =(x1, x2, x3, ..., x T ), T represents the total number of data in the first intermediate data sequence; S3: Set the number of iterations, and convert the first intermediate data sequence D T =(x1, x2, x3, ..., x T ) is input into the iterative model for iteration to obtain the predicted data at time n, where n is less than or equal to T; S4: Compare the predicted data at time n with a preset data threshold, and adjust the priority of the patent data scheduling task and control the scheduling task of each computing resource according to the comparison result to realize patent data scheduling.

2. The patent data scheduling method according to claim 1, characterized in that: In step S2, the historical CPU usage data and the historical memory usage data are preprocessed to obtain a first intermediate data sequence D T =(x1, x2, x3, ..., x T ), specifically: S21: performing differentiation and filtering processing on the historical CPU usage data and the historical memory usage data to obtain processed historical CPU usage data and processed historical memory usage data; S22: intercepting the first proportion of processed historical CPU usage data; S23: intercepting the processed historical memory usage data of the second proportion; S24: Combine the first-ratio processed historical CPU usage data obtained in step S22 and the second-ratio processed historical memory usage data obtained in step S23 as the first intermediate data sequence D T =(x1, x2, x3, ..., x T ).

3. The patent data scheduling method as claimed in claim 2, characterized in that: The first ratio is 80%, and the second ratio is 70%.

4. The patent data scheduling method according to claim 2, characterized in that: In the step S24, the processed historical CPU usage data of the first ratio obtained in step S22 and the processed historical memory usage data of the second ratio obtained in step S23 are combined, specifically, the processed historical CPU usage data of the first ratio obtained in step S22 is first, and the processed historical memory usage data of the second ratio obtained in step S23 is later, and the data are spliced ​​to obtain the first intermediate data sequence D T .

5. The patent data scheduling method according to claim 1, characterized in that: The iterative model in step S3 is specifically: s t-1 =k1×random(0,1) ITE t-1 =k2×random(0,1); Among them, LF t represents the logistic equation at time t, which is used as the threshold function of the iterative model. e represents a natural constant, avg() represents an average function, sinh() represents a hyperbolic sine function, cosh() represents a hyperbolic cosine function, random() represents a random number function, and x t represents the data at time t in the first intermediate data sequence, t = 2, 3, ..., n, s t and ITE t Respectively represent the input data state variable and the iteration state variable at time t, s1 and ITE1 represent the initial value of the input data state variable and the initial value of the iteration state variable respectively, the input data state variable at time t is used for data cleaning after each iteration, the iteration state variable at time t is used to characterize the output data after each iteration, k1 and k2 represent the first adjustment coefficient and the second adjustment coefficient respectively. .

6. The patent data scheduling method as claimed in claim 5, characterized in that: In step S4, the predicted data at time n is compared with the preset data threshold, and the priority of the patent data scheduling task of each computing resource is adjusted and the scheduling task is controlled according to the comparison result. Specifically, the iteration state variable ITE is obtained by iteration. n , and the iteration state variable ITE n The data at the predicted time n is compared with the preset data threshold. When the comparison result is greater than or equal to, the priority of the current patent data scheduling task is reduced and the current task execution is suspended; when the comparison result is less than, the priority of the current patent data scheduling task is kept unchanged.

7. A patent data scheduling system, comprising a storage subsystem, a monitoring subsystem and a data fusion subsystem, wherein the data fusion subsystem is used to integrate, clean and fuse patent data from different data sources, and send the processed patent data to the storage subsystem for storage of the patent data, and the monitoring subsystem uses multi-threading technology to divide each computing resource by IP, and performs polling monitoring and control on the patent data scheduling tasks of each computing resource, characterized in that: The monitoring subsystem adjusts the priority of patent data scheduling tasks and controls scheduling tasks for each computing resource according to the patent data scheduling method as described in any one of claims 1 to 6 to implement patent data scheduling.

Images