Decision-making system and method based on big data optimization analysis

Abstract

The invention discloses a decision-making system and method based on big data optimization analysis, and the system comprises a data obtaining module which is used for obtaining the first production process data of an enterprise, and crawling the second production process data of a benchmarking enterprise corresponding to the enterprise; the preprocessing module is used for performing denoising, deduplication and standardization processing on the first production process data and the second production process data to obtain first target data and second target data; the analysis module is used for performing clustering analysis on the second target data, dividing the second target data into a plurality of data group classes, and determining a target data group class matched with the first target data in each data group class; and the decision and suggestion module is used for determining an adjacent data group class corresponding to the target data group class, and generating decision and suggestion information corresponding to the production process of the enterprise based on the adjacent data group class so as to optimize the production process of the enterprise based on the decision and suggestion information. According to the invention, efficient and accurate analysis processing of the big data related to the enterprise is realized.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a decision-making system and method based on big data optimization analysis. Background Art

[0002] With the rapid development of information technology, big data has become an important resource for enterprises, and enterprises can use big data to promote their development. However, the massive and diverse characteristics of big data have brought great challenges to data analysis and processing. Traditional data analysis and processing methods have problems such as low efficiency and insufficient accuracy when processing big data, and cannot meet the needs of modern enterprises for efficient use of data. Therefore, how to efficiently and accurately analyze and process big data related to enterprises is one of the technical problems that need to be solved. Summary of the invention

[0003] The main purpose of the present invention is to provide a decision-making system and method based on big data optimization analysis, aiming to solve the technical problem of how to efficiently and accurately analyze and process big data related to enterprises in the prior art.

[0004] To achieve the above object, the present invention provides a decision-making system based on big data optimization analysis, and the decision-making system based on big data optimization analysis includes:

[0005] A data acquisition module, used to acquire first production process data of an enterprise, and crawl second production process data of a benchmark enterprise corresponding to the enterprise;

[0006] A preprocessing module, used for performing denoising, deduplication and standardization processing on the first production process data and the second production process data respectively to obtain first target data and second target data;

[0007] an analysis module, configured to perform cluster analysis on the second target data, divide the second target data into a plurality of data group classes, and determine a target data group class that matches the first target data in each of the data group classes;

[0008] A decision suggestion module is used to determine the adjacent data group class corresponding to the target data group class, and generate decision suggestion information corresponding to the production process of the enterprise based on the adjacent data group class, so as to optimize the production process of the enterprise based on the decision suggestion information.

[0009] Preferably, the decision suggestion module includes:

[0010] A judging unit, configured to judge whether a first production efficiency corresponding to the adjacent data group class is higher than a second production efficiency corresponding to the target data group class;

[0011] A search unit, configured to search for reference benchmarking enterprises from which each element in the adjacent data group class originates, and search for reference production process data corresponding to each of the reference benchmarking enterprises, when the first production efficiency is higher than the second production efficiency;

[0012] A generating unit is used to generate the decision suggestion information based on the reference production process data and the first production process data.

[0013] Preferably, the analysis module comprises:

[0014] A comparison unit, used for comparing the enterprise production characteristics between the second production process data corresponding to each of the data groups and the first production process data, to obtain a plurality of comparison results, wherein the enterprise production characteristics at least include production scale characteristics, production type characteristics, production process characteristics and production efficiency characteristics;

[0015] The first determination unit is used to determine a target comparison result among the comparison results, and determine the data group class corresponding to the target comparison result as the target data group class, wherein the target comparison result is a comparison result of a successful match of enterprise production characteristics between the second production process data and the first production process data.

[0016] Preferably, the analysis module further includes:

[0017] a second determining unit, configured to randomly determine a plurality of second target data from the second target data as initial cluster centers, and use other data in the second target data except the initial cluster centers as data to be clustered;

[0018] A calculation unit is used to calculate the distance between each of the to-be-clustered data and each of the initial cluster centers, and generate an initial cluster group according to the distance. The calculation formula of the distance is:

[0019]

[0020] Where d represents the distance, n represents the number of initial cluster centers, x represents the data to be clustered, and ci represents the i-th initial cluster center;

[0021] The second determination unit is further used to determine the group cluster center of each of the initial cluster groups, and determine whether the group cluster center is the same as the initial cluster center, and if they are the same, generate each of the initial cluster groups into a plurality of the data group classes;

[0022] An execution unit is used for, if the group cluster center is different from the initial cluster center, taking the group cluster center as a new initial cluster center, and executing the step of calculating the distance between each of the to-be-clustered data and each of the initial cluster centers.

[0023] Preferably, the decision-making system further includes:

[0024] A storage module is used to store the decision suggestion information, the adjacent data group class and the first production process data.

[0025] Furthermore, to achieve the above-mentioned purpose, the present invention also provides a decision-making method based on big data optimization analysis, and the decision-making method based on big data optimization analysis includes:

[0026] Acquire the first production process data of the enterprise, and crawl the second production process data of the benchmark enterprise corresponding to the enterprise;

[0027] Performing denoising, deduplication and standardization processing on the first production process data and the second production process data respectively to obtain first target data and second target data;

[0028] Performing cluster analysis on the second target data based on a preset analysis model, dividing the second target data into a plurality of data group classes, and determining a target data group class matching the first target data in each of the data group classes;

[0029] Determine a neighboring data group class corresponding to the target data group class, and generate decision suggestion information corresponding to the production process of the enterprise based on the neighboring data group class, so as to optimize the production process of the enterprise based on the decision suggestion information.

[0030] Preferably, the step of generating decision suggestion information corresponding to the production process of the enterprise based on the adjacent data group class includes:

[0031] Determining whether a first production efficiency corresponding to the adjacent data group class is higher than a second production efficiency corresponding to the target data group class;

[0032] If the first production efficiency is higher than the second production efficiency, then searching for reference benchmarking enterprises from which each element in the adjacent data group class originates, and searching for reference production process data corresponding to each of the reference benchmarking enterprises;

[0033] The decision suggestion information is generated based on the reference production process data and the first production process data.

[0034] Preferably, the step of determining the target data group class matching the first target data in each of the data group classes comprises:

[0035] Comparing the enterprise production characteristics between the second production process data corresponding to each of the data groups and the first production process data to obtain a plurality of comparison results, wherein the enterprise production characteristics at least include production scale characteristics, production type characteristics, production process characteristics and production efficiency characteristics;

[0036] Determine a target comparison result among the comparison results, and determine the data group class corresponding to the target comparison result as the target data group class, wherein the target comparison result is a comparison result in which the enterprise production characteristics between the second production process data and the first production process data are successfully matched.

[0037] Preferably, the step of performing cluster analysis on the second target data based on a preset analysis model to divide the second target data into a plurality of data groups comprises:

[0038] Randomly determine a plurality of second target data from the second target data as initial cluster centers, and use other data in the second target data except the initial cluster centers as data to be clustered;

[0039] The distance between each of the data to be clustered and each of the initial cluster centers is calculated, and an initial cluster group is generated according to the distance. The calculation formula of the distance is:

[0040]

[0041] Where d represents the distance, n represents the number of initial cluster centers, x represents the data to be clustered, and ci represents the i-th initial cluster center;

[0042] Determine the group cluster center of each of the initial cluster groups, and determine whether the group cluster center is the same as the initial cluster center, and if they are the same, generate each of the initial cluster groups into a plurality of the data group classes;

[0043] If the group cluster center is different from the initial cluster center, the group cluster center is used as a new initial cluster center, and the step of calculating the distance from each of the to-be-clustered data to each of the initial cluster centers is performed.

[0044] Preferably, after the step of generating decision suggestion information corresponding to the production process of the enterprise based on the adjacent data group class, the step further includes:

[0045] The decision suggestion information, the neighboring data group class and the first production process data are stored.

[0046] The present invention provides a decision-making system and method based on big data optimization analysis. The decision-making system based on big data optimization analysis includes a data acquisition module, a preprocessing module, an analysis module and a decision suggestion module; wherein the data acquisition module is used to acquire the first production process data of the enterprise, and crawl the second production process data of the benchmark enterprise corresponding to the enterprise; the preprocessing module is used to perform denoising, deduplication and standardization on the first production process data and the second production process data, respectively, to obtain the first target data and the second target data; the analysis module is used to perform cluster analysis on the second target data, divide the second target data into multiple data group classes, and determine the target data group class in each data group class that matches the first target data; the decision suggestion module is used to determine the adjacent data group class corresponding to the target data group class, and generate decision suggestion information corresponding to the production process of the enterprise based on the adjacent data group class, so as to optimize the production process of the enterprise based on the decision suggestion information. In this way, the data of the benchmark enterprises corresponding to the enterprise in each production process are collected with the help of crawler technology for denoising, deduplication and standardization preprocessing, as well as cluster analysis, to obtain multiple data groups, and determine the target data group to which the data of the enterprise itself in each production process belongs, and then determine the adjacent data group that represents a higher production efficiency than the target data group, and generate decision-making recommendation information for enterprise decision makers to refer to the decision-making recommendation information to optimize the production process of the enterprise and improve the production efficiency of the enterprise. Among them, the interference of invalid data is reduced through preprocessing, and the efficiency of data processing is improved, while cluster analysis can discover the potential associations and laws between data, making the analysis more accurate, and realizing efficient and accurate analysis and processing of big data related to the enterprise. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 This is a module diagram of the first embodiment of the decision-making system based on big data optimization analysis of the present invention;

[0048] Figure 2 This is a module diagram of a second embodiment of a decision-making system based on big data optimization analysis of the present invention;

[0049] Figure 3 It is a flowchart of the first embodiment of the decision-making method based on big data optimization analysis of the present invention;

[0050] Figure 4 It is a flow chart of the second embodiment of the decision-making method based on big data optimization analysis of the present invention.

[0051] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0052] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0053] The present invention provides a decision-making system based on big data optimization analysis. Figure 1 , Figure 1 This is a module diagram of the first embodiment of the decision-making system based on big data optimization analysis of the present invention.

[0054] Specifically, the decision-making system based on big data optimization analysis in this embodiment includes a data acquisition module 10, a preprocessing module 20, an analysis module 30 and a decision suggestion module 40, wherein:

[0055] The data acquisition module 10 is used to acquire the first production process data of the enterprise and crawl the second production process data of the benchmark enterprise corresponding to the enterprise;

[0056] Furthermore, when an enterprise needs to make a decision through big data optimization analysis, it initiates a reference decision request to the decision system based on big data optimization analysis of this embodiment. After receiving the reference decision request, the decision system reads the enterprise ID (Identification) carried therein, and determines the specific enterprise that needs to perform big data optimization analysis and generate relevant reference decisions based on the analysis by the enterprise ID.

[0057] Furthermore, the decision-making system initiates a big data optimization analysis for the enterprise, and the data acquisition module therein acquires the first production process data of the enterprise. The first production process data is the data of the production process of the enterprise itself that needs to be analyzed, including but not limited to daily production time, daily output, daily production personnel, production tool type, production process, etc. This type of data can be uploaded by the enterprise itself and received by the acquisition module, or it can be formed by the acquisition module crawling relevant data from the Internet, or it can be uploaded by the enterprise itself and crawled.

[0058] In addition, the acquisition module also crawls the second production process data of the benchmarking enterprises corresponding to the enterprise from various official websites, news websites, social platforms and other channels by calling API (Application Programming Interface), web crawlers and other technologies. Among them, the benchmarking enterprise is an enterprise that belongs to the same industry as the enterprise and has a competitive relationship with the enterprise, and the second production process data is various types of data related to the production process of the benchmarking enterprise.

[0059] A preprocessing module 20, configured to perform denoising, deduplication and standardization processing on the first production process data and the second production process data, respectively, to obtain first target data and second target data;

[0060] Understandably, the first production process data and the second production process data may be crawled from the Internet, and there may be data irrelevant to the actual situation of the enterprise, repeated data, and data with the same meaning but different numerical units, and there may also be data in the form of graphics. In this regard, preprocessing is performed through a preprocessing module, and the preprocessing includes at least denoising, deduplication and standardized conversion processing.

[0061] Specifically, the data irrelevant to the enterprise in the first production process data and the second production process data are searched as noise data, and the noise data is removed from the two for denoising. At the same time, the duplicate data in the two are identified, and any one item is selected from each duplicate data to be marked as a retained item, and the other data is marked as a deleted item, and then the data marked as the deleted item is deleted from the two for deduplication. For the data in the two that have the same meaning but different numerical units, they can be converted according to a preset unified numerical unit. For the graphics in the two, the text therein is identified to obtain the text information of the graphics in the two. In this way, the preprocessing of the first production process data and the second production process data is realized by removing the noise data, deleting the duplicate items, converting the numerical units, and identifying the text information in the graphics, and the first target data and the second target data corresponding to the two are obtained by preprocessing, which reduces the interference of invalid data and improves the efficiency of subsequent data processing.

[0062] An analysis module 30, configured to perform cluster analysis on the second target data, divide the second target data into a plurality of data group classes, and determine a target data group class that matches the first target data in each of the data group classes;

[0063] Furthermore, the analysis module performs cluster analysis on the second target data obtained through preprocessing, discovers potential associations and rules between data through cluster analysis, and then divides data with the same characteristics into the same data group class according to the associations and rules of each data in the second target data, so as to obtain multiple data group classes with more accurate division. Then, the first target data is compared with each data group class, and the data group class matching the first target data is determined as the target data group class. Among them, whether it matches or not can be determined by data characteristics. Specifically, the analysis module includes:

[0064] A comparison unit 31 is used to compare the enterprise production characteristics between the second production process data corresponding to each of the data groups and the first production process data to obtain multiple comparison results, wherein the enterprise production characteristics at least include production scale characteristics, production type characteristics, production process characteristics and production efficiency characteristics;

[0065] The first determination unit 32 is used to determine a target comparison result among the comparison results, and determine the data group class corresponding to the target comparison result as the target data group class, wherein the target comparison result is a comparison result in which the enterprise production characteristics between the second production process data and the first production process data are successfully matched.

[0066] Furthermore, for each data group class, the comparison unit compares the corresponding second production process data, i.e., the enterprise production characteristics represented by the second production process data assigned to the data group class, with the enterprise production characteristics represented by the first production process data to obtain a corresponding comparison result. Among them, the enterprise generation characteristics at least include the enterprise's production scale characteristics, such as the number of daily production personnel, daily production time, daily production volume, etc., generation type characteristics, such as fully automated production, semi-automated production, manual production, etc., production process, such as production includes 5 production steps, or includes 4 steps, or the order between the steps is different, etc., and production efficiency characteristics, etc.

[0067] Furthermore, after each data group class is compared with the first production process data, multiple comparison results are generated. The multiple comparison results can be divided into two categories. One is that the enterprise production characteristics represented by the second production process data corresponding to the data group class match the enterprise characteristics represented by the first production process data, that is, the two have similarities in production scale characteristics, production type characteristics, production process characteristics and production efficiency characteristics; the other is that the enterprise production characteristics represented by the second production process data corresponding to the data group class do not match the enterprise characteristics represented by the first production process data, that is, the two have no similarity in at least one of the production scale characteristics, production type characteristics, production process characteristics and production efficiency characteristics.

[0068] In addition, the comparison results of feature mismatch can be divided into two categories. One is that the enterprise production characteristics represented by the second production process data corresponding to the data group class are lower in production efficiency than the enterprise characteristics represented by the first production process data; the other is that the enterprise production characteristics represented by the second production process data corresponding to the data group class are higher in production efficiency than the enterprise characteristics represented by the first production process data. The production process of the benchmarking enterprise with this type of second production process data can be used as a reference for the enterprise.

[0069] Furthermore, the first determination unit searches for a matching comparison result in each comparison result as a target comparison result, and then determines a data group class generating the target comparison result, and determines the data group class as the target data group class. The second production process data in the target data group class has similar enterprise production characteristics to the first production process data.

[0070] The decision suggestion module 40 is used to determine the neighboring data group class corresponding to the target data group class, and generate decision suggestion information corresponding to the production process of the enterprise based on the neighboring data group class, so as to optimize the production process of the enterprise based on the decision suggestion information.

[0071] Furthermore, the decision suggestion module determines the neighboring data group class corresponding to the target data group class based on each comparison. The neighboring data group class is for the data group class that does not match the enterprise characteristics represented by the first production process data, and is arranged in order of production efficiency from low to high, and is the data group class before and after the target data group class. Then, decision suggestion information corresponding to the enterprise production process is generated based on the neighboring data group class for reference by the enterprise decision makers, and the enterprise's production process is optimized based on the decision suggestion information to improve the enterprise's production efficiency. Because the neighboring data group class includes a data group class that is ranked before the target data group class, indicating that the production efficiency is lower than the target data group class, and also includes a data group class that is ranked after the target data group class, indicating that the production efficiency is higher than the target data group class, it is necessary to distinguish when generating decision suggestion information based on it. Specifically, the decision suggestion module includes:

[0072] A judging unit 41, configured to judge whether the first production efficiency corresponding to the adjacent data group is higher than the second production efficiency corresponding to the target data group;

[0073] A search unit 42 is used to search for reference benchmarking enterprises from which each element in the adjacent data group class originates when the first production efficiency is higher than the second production efficiency, and to search for reference production process data corresponding to each of the reference benchmarking enterprises;

[0074] The generating unit 43 is used to generate the decision suggestion information based on the reference production process data and the first production process data.

[0075] Furthermore, the production efficiency represented by the adjacent data group class is taken as the first production efficiency, and the production efficiency corresponding to the target data group class is taken as the second production efficiency, and the judgment unit is used to judge whether the first production efficiency is higher than the second production efficiency. If the first production efficiency is greater than the second production efficiency, it means that the adjacent data group class is a data group class arranged after the target data group class, and the production process of the corresponding benchmarking enterprise can be used as a reference for the enterprise production process. At this time, the second generated process data forming the adjacent data group class is used as an element of the adjacent data group class, and the benchmarking enterprise from which each element is derived is searched by the search unit as a reference benchmarking enterprise, and the production process data corresponding to each reference benchmarking enterprise is searched as the reference production process data. Then, the generation unit generates decision suggestion information based on the difference between the reference production process data and the first production process data. By viewing the decision suggestion information, the enterprise decision maker can know the difference between the production process data of his enterprise and the reference benchmarking enterprise, and then optimize the enterprise's production process based on the difference to improve the enterprise's production efficiency.

[0076] It should be noted that the decision system may also be provided with a storage module 50, through which the generated decision suggestion information, the adjacent data group class used to generate the decision suggestion and the first production process data of the enterprise itself are stored, so as to facilitate subsequent review and tracing.

[0077] The decision-making system based on big data optimization analysis implemented in this embodiment includes a data acquisition module, a preprocessing module, an analysis module and a decision suggestion module; wherein the data acquisition module is used to acquire the first production process data of the enterprise, and crawl the second production process data of the benchmark enterprise corresponding to the enterprise; the preprocessing module is used to perform denoising, deduplication and standardization on the first production process data and the second production process data respectively, to obtain the first target data and the second target data; the analysis module is used to perform cluster analysis on the second target data, divide the second target data into multiple data group classes, and determine the target data group class in each data group class that matches the first target data; the decision suggestion module is used to determine the adjacent data group class corresponding to the target data group class, and generate decision suggestion information corresponding to the production process of the enterprise based on the adjacent data group class, so as to optimize the production process of the enterprise based on the decision suggestion information. In this way, the data of the benchmark enterprises corresponding to the enterprise in each production process are collected with the help of crawler technology for denoising, deduplication and standardization preprocessing, as well as cluster analysis, to obtain multiple data groups, and determine the target data group to which the data of the enterprise itself in each production process belongs, and then determine the adjacent data group that represents a higher production efficiency than the target data group, and generate decision-making recommendation information for enterprise decision makers to refer to the decision-making recommendation information to optimize the production process of the enterprise and improve the production efficiency of the enterprise. Among them, the interference of invalid data is reduced through preprocessing, and the efficiency of data processing is improved, while cluster analysis can discover the potential associations and laws between data, making the analysis more accurate, and realizing efficient and accurate analysis and processing of big data related to the enterprise.

[0078] For further information, please refer to Figure 2 Based on the first embodiment of the decision-making system based on big data optimization analysis of the present invention, a second embodiment of the decision-making system based on big data optimization analysis of the present invention is proposed.

[0079] The difference between the second embodiment of the decision-making system based on big data optimization analysis and the first embodiment of the decision-making system based on big data optimization analysis is that the analysis module further includes:

[0080] A second determining unit 33 is used to randomly determine a plurality of second target data from the second target data as initial cluster centers, and use other data in the second target data except the initial cluster centers as data to be clustered;

[0081] A calculation unit 34, used for calculating the distance between each of the to-be-clustered data and each of the initial cluster centers, and generating an initial cluster group according to the distance;

[0082] The second determination unit 33 is further used to determine the group cluster center of each of the initial cluster groups, and determine whether the group cluster center is the same as the initial cluster center, and if they are the same, generate each of the initial cluster groups into a plurality of the data group classes;

[0083] The execution unit 35 is used for, if the group cluster center is different from the initial cluster center, taking the group cluster center as a new initial cluster center, and executing the step of calculating the distance between each of the to-be-clustered data and each of the initial cluster centers.

[0084] Furthermore, in order to perform accurate cluster analysis on the second target data, the present embodiment divides the analysis module into a second determination unit, a calculation unit, and an execution unit. First, the second determination unit randomly determines a plurality of second target data from the second target data as initial cluster centers. The specific number of initial cluster centers can be preset, for example, if it is set to 5, then 5 second target data are randomly selected as initial cluster centers. Then, the other data in the second target data except the initial cluster centers are used as data to be clustered, and the distance between each data to be clustered and each initial cluster center is calculated by the calculation unit. The specific calculation formula can be found in the following formula (1).

[0085]

[0086] Among them, d represents the distance, n represents the number of initial clustering centers, x represents the data to be clustered, which can be the vector mean of the production scale characteristics, production type characteristics, production process characteristics and production efficiency characteristics represented by the data to be clustered, and ci represents the i-th initial clustering center, which can be the vector mean of the production scale characteristics, production type characteristics, production process characteristics and production efficiency characteristics represented by the initial distance center.

[0087] Furthermore, after calculating the distance from each data to be clustered to each initial cluster center, each distance is compared to determine the minimum distance, and the data to be clustered is divided into the initial cluster center corresponding to the minimum value. After each data to be clustered is divided into the initial cluster center corresponding to the respective minimum value, a plurality of initial cluster groups are obtained. The number of the initial cluster groups is consistent with the number of cluster centers.

[0088] Further, for each initial clustering group, the second determination unit also performs mean operation to obtain a mean result, and then calculates the distance between each data in the initial clustering group and the mean result, and selects the shortest distance to be determined as the group clustering center of the initial clustering group. After each initial clustering group determines its own group clustering center, the group clustering center of each initial clustering group is compared with its initial clustering center to determine whether the group clustering center of each initial clustering group is the same as its initial clustering center. If they are the same, it means that the division of the initial clustering center and the initial clustering group is accurate, so each initial clustering group is generated into multiple data groups. On the contrary, if there is any initial clustering group whose group clustering center is different from its initial clustering center, the group clustering center is updated to a new initial clustering center, and other data are used as new data to be clustered. Then, for the new initial cluster center and the new data to be clustered, the distance from each data to be clustered to each initial cluster center is calculated, and this cycle is repeated until the cluster center no longer changes, and the group cluster center of each initial cluster group is the same as its initial cluster center, and each initial cluster group is generated into multiple data groups.

[0089] This implementation performs a cyclic clustering analysis on the second target data and classifies the second target data with similar characteristics into the same group according to the characteristics of the second target data in production scale, production type, production process and production efficiency, thereby achieving accurate classification of the second target data.

[0090] In addition, the present invention provides a decision-making method based on big data optimization analysis, please refer to Figure 3 , Figure 3 This is a flowchart of a first embodiment of the decision-making method based on big data optimization analysis of the present invention. It should be noted that although a logical sequence is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than that shown here.

[0091] Specifically, the decision-making method based on big data optimization analysis in this embodiment includes:

[0092] Step S10, obtaining first production process data of an enterprise, and crawling second production process data of a benchmark enterprise corresponding to the enterprise;

[0093] Furthermore, when an enterprise needs to make a decision through big data optimization analysis, it initiates a reference decision request to the decision system based on big data optimization analysis of this embodiment. After receiving the reference decision request, the decision system reads the enterprise ID (Identification) carried therein, and determines the specific enterprise that needs to perform big data optimization analysis and generate relevant reference decisions based on the analysis by the enterprise ID.

[0094] Furthermore, the decision system starts big data optimization analysis for the enterprise and acquires the first production process data of the enterprise. The first production process data is the data of the enterprise's own production process that needs to be analyzed, including but not limited to daily production time, daily output, daily production personnel, production tool type, production process, etc. This type of data can be uploaded by the enterprise itself and received by the acquisition module, or it can be formed by the acquisition module crawling relevant data from the Internet, or it can be uploaded by the enterprise itself and crawled.

[0095] In addition, the second production process data of the benchmarking enterprises corresponding to the enterprise are crawled from various official websites, news websites, social platforms and other channels by calling API (Application Programming Interface), web crawlers and other technologies. Among them, the benchmarking enterprise is an enterprise that belongs to the same industry as the enterprise and has a competitive relationship with the enterprise, and the second production process data is various types of data related to the production process of the benchmarking enterprise.

[0096] Step S20, performing denoising, deduplication and standardization processing on the first production process data and the second production process data respectively to obtain first target data and second target data;

[0097] Understandably, the first production process data and the second production process data may both be crawled from the Internet, and may contain data irrelevant to the actual situation of the enterprise, repeated data, and data with the same meaning but different numerical units, and data in the form of graphics. Therefore, the first production process data and the second production process data are preprocessed respectively, and the preprocessing includes at least denoising, deduplication and standardized conversion processing.

[0098] Specifically, the data irrelevant to the enterprise in the first production process data and the second production process data are searched as noise data, and the noise data is removed from the two for denoising. At the same time, the duplicate data in the two are identified, and any one item is selected from each duplicate data to be marked as a retained item, and the other data is marked as a deleted item, and then the data marked as the deleted item is deleted from the two for deduplication. For the data in the two that have the same meaning but different numerical units, they can be converted according to a preset unified numerical unit. For the graphics in the two, the text therein is identified to obtain the text information of the graphics in the two. In this way, the preprocessing of the first production process data and the second production process data is realized by removing the noise data, deleting the duplicate items, converting the numerical units, and identifying the text information in the graphics, and the first target data and the second target data corresponding to the two are obtained by preprocessing, which reduces the interference of invalid data and improves the efficiency of subsequent data processing.

[0099] Step S30, performing cluster analysis on the second target data, dividing the second target data into a plurality of data group classes, and determining a target data group class matching the first target data in each of the data group classes;

[0100] Furthermore, cluster analysis is performed on the second target data obtained through preprocessing, and potential associations and rules between data are discovered through cluster analysis. Then, according to the associations and rules of each data in the second target data, data with the same characteristics are divided into the same data group class, and multiple data group classes with more accurate divisions are obtained. Then, the first target data is compared with each data group class, and the data group class that matches the first target data is determined as the target data group class. Among them, whether it matches or not can be determined by data characteristics. Specifically, the step of determining the target data group class that matches the first target data in each of the data groups includes:

[0101] Step S31, comparing the enterprise production characteristics between the second production process data corresponding to each of the data groups and the first production process data to obtain multiple comparison results, wherein the enterprise production characteristics at least include production scale characteristics, production type characteristics, production process characteristics and production efficiency characteristics;

[0102] Step S32, determining a target comparison result among the comparison results, and determining the data group class corresponding to the target comparison result as the target data group class, wherein the target comparison result is a comparison result of a successful match between the enterprise production characteristics of the second production process data and the first production process data.

[0103] Furthermore, for each data group class, the corresponding second production process data, that is, the enterprise production characteristics represented by the second production process data assigned to the data group class, are compared with the enterprise production characteristics represented by the first production process data to obtain corresponding comparison results. Among them, the enterprise generation characteristics at least include the production scale characteristics of the enterprise, such as the number of daily production personnel, daily production time, daily production volume, etc., generation type characteristics, such as fully automated production, semi-automated production, manual production, etc., production process, such as production includes 5 production steps, or includes 4 steps, or the order between the steps is different, etc., and production efficiency characteristics, etc.

[0104] Furthermore, after each data group class is compared with the first production process data, multiple comparison results are generated. The multiple comparison results can be divided into two categories. One is that the enterprise production characteristics represented by the second production process data corresponding to the data group class match the enterprise characteristics represented by the first production process data, that is, the two have similarities in production scale characteristics, production type characteristics, production process characteristics and production efficiency characteristics; the other is that the enterprise production characteristics represented by the second production process data corresponding to the data group class do not match the enterprise characteristics represented by the first production process data, that is, the two have no similarity in at least one of the production scale characteristics, production type characteristics, production process characteristics and production efficiency characteristics.

[0105] In addition, the comparison results of feature mismatch can be divided into two categories. One is that the enterprise production characteristics represented by the second production process data corresponding to the data group class are lower in production efficiency than the enterprise characteristics represented by the first production process data; the other is that the enterprise production characteristics represented by the second production process data corresponding to the data group class are higher in production efficiency than the enterprise characteristics represented by the first production process data. The production process of the benchmarking enterprise with this type of second production process data can be used as a reference for the enterprise.

[0106] Furthermore, a matching comparison result in each comparison result is searched as a target comparison result, and then a data group class generating the target comparison result is determined, and the data group class is determined as the target data group class. The second production process data in the target data group class has similar enterprise production characteristics to the first production process data.

[0107] Step S40, determining a neighboring data group class corresponding to the target data group class, and generating decision suggestion information corresponding to the production process of the enterprise based on the neighboring data group class, so as to optimize the production process of the enterprise based on the decision suggestion information.

[0108] Furthermore, a neighboring data group class corresponding to the target data group class is determined based on each comparison. The neighboring data group class is a data group class that does not match the enterprise characteristics represented by the first production process data, and is arranged in order of production efficiency from low to high, and is a data group class that is arranged before and after the target data group class. Then, decision-making recommendation information corresponding to the enterprise's production process is generated based on the neighboring data group class for reference by the enterprise's decision makers, and the enterprise's production process is optimized based on the decision-making recommendation information to improve the enterprise's production efficiency. Because the neighboring data group class includes a data group class that is ranked before the target data group class, indicating that the production efficiency is lower than the target data group class, and also includes a data group class that is ranked after the target data group class, indicating that the production efficiency is higher than the target data group class, it is necessary to distinguish when generating decision-making recommendation information based on it. Specifically, the step of generating decision-making recommendation information corresponding to the production process of the enterprise based on the neighboring data group class includes:

[0109] Step S41, determining whether the first production efficiency corresponding to the adjacent data group class is higher than the second production efficiency corresponding to the target data group class;

[0110] Step S42, if the first production efficiency is higher than the second production efficiency, searching for reference benchmarking enterprises from which each element in the adjacent data group class originates, and searching for reference production process data corresponding to each of the reference benchmarking enterprises;

[0111] Step S43: generating the decision suggestion information based on the reference production process data and the first production process data.

[0112] Furthermore, the production efficiency represented by the adjacent data group class is taken as the first production efficiency, and the production efficiency corresponding to the target data group class is taken as the second production efficiency, and it is determined whether the first production efficiency is higher than the second production efficiency. If the first production efficiency is greater than the second production efficiency, it means that the adjacent data group class is a data group class arranged after the target data group class, and the production process of the corresponding benchmarking enterprise can be used as a reference for the enterprise production process. At this time, the second generated process data forming the adjacent data group class is used as an element of the adjacent data group class, and the benchmarking enterprise from which each element is derived is searched through the search unit as a reference benchmarking enterprise, and the production process data corresponding to each reference benchmarking enterprise is searched as the reference production process data. Then, based on the difference between the reference production process data and the first production process data, decision-making recommendation information is generated. By viewing the decision-making recommendation information, the decision maker of the enterprise can know the difference between the production process data of his enterprise and the reference benchmarking enterprise, and then optimize the production process of the enterprise based on the difference to improve the production efficiency of the enterprise.

[0113] It should be noted that the present implementation can also store the generated decision recommendation information, that is, after the decision recommendation information is generated, the generated decision recommendation information, the adjacent data group class used to generate the decision recommendation and the first production process data of the enterprise itself are stored to facilitate subsequent review and tracing.

[0114] The decision-making method based on big data optimization analysis implemented in this implementation first obtains the first production process data of the enterprise, and crawls the second production process data of the benchmark enterprise corresponding to the enterprise; and performs denoising, deduplication and standardization on the first production process data and the second production process data respectively, to obtain the first target data and the second target data; and then performs cluster analysis on the second target data, divides the second target data into multiple data groups, and determines the target data group in each data group that matches the first target data; and determines the adjacent data group corresponding to the target data group, and generates decision suggestion information corresponding to the production process of the enterprise based on the adjacent data group, so as to optimize the production process of the enterprise based on the decision suggestion information. In this way, with the help of crawler technology, the data of the benchmark enterprise corresponding to the enterprise in each production process is collected for denoising, deduplication and standardization preprocessing, and cluster analysis is performed to obtain multiple data groups, and the target data group to which the data of the enterprise itself in each production process belongs is determined, and then the adjacent data group representing a higher production efficiency than the target data group is determined, and decision suggestion information is generated, so that the decision makers of the enterprise can refer to the decision suggestion information to optimize the production process of the enterprise and improve the production efficiency of the enterprise. Among them, preprocessing reduces the interference of invalid data and improves the efficiency of data processing, while cluster analysis can discover the potential connections and patterns between data, making the analysis more accurate and realizing efficient and accurate analysis and processing of big data related to the enterprise.

[0115] For further information, please refer to Figure 4 Based on the first embodiment of the decision-making method based on big data optimization analysis of the present invention, a second embodiment of the decision-making method based on big data optimization analysis of the present invention is proposed.

[0116] The difference between the second embodiment of the decision-making method based on big data optimization analysis and the first embodiment of the decision-making method based on big data optimization analysis is that the step of performing cluster analysis on the second target data based on a preset analysis model and dividing the second target data into a plurality of data groups includes:

[0117] Step S33, randomly determining a plurality of second target data from the second target data as initial cluster centers, and taking other data in the second target data except the initial cluster centers as data to be clustered;

[0118] Step S34, calculating the distance between each of the to-be-clustered data and each of the initial cluster centers, and generating an initial cluster group according to the distance;

[0119] Step S35, determining the group cluster center of each of the initial cluster groups, and judging whether the group cluster center is the same as the initial cluster center, and if they are the same, generating each of the initial cluster groups into a plurality of the data group classes;

[0120] Step S36: if the group cluster center is different from the initial cluster center, the group cluster center is used as a new initial cluster center, and the step of calculating the distance from each of the to-be-clustered data to each of the initial cluster centers is performed.

[0121] Furthermore, a plurality of second target data are randomly determined from the second target data as initial cluster centers. The specific number of initial cluster centers can be preset, for example, if it is set to 5, then 5 second target data are randomly selected as initial cluster centers. Then, the other data in the second target data except the initial cluster centers are used as the data to be clustered, and the distance between each data to be clustered and each initial cluster center is calculated. The specific calculation formula can be referred to as the above formula (1).

[0122] Furthermore, after calculating the distance from each data to be clustered to each initial cluster center, each distance is compared to determine the minimum distance, and the data to be clustered is divided into the initial cluster center corresponding to the minimum value. After each data to be clustered is divided into the initial cluster center corresponding to the respective minimum value, a plurality of initial cluster groups are obtained. The number of the initial cluster groups is consistent with the number of cluster centers.

[0123] Further, a mean operation is performed for each initial clustering group to obtain a mean result, and then the distance between each data in the initial clustering group and the mean result is calculated, and the shortest distance is selected to determine the group cluster center of the initial clustering group. After each initial clustering group determines its own group cluster center, the group cluster center of each initial clustering group is compared with its initial cluster center to determine whether the group cluster center of each initial clustering group is the same as its initial cluster center. If they are the same, it means that the division of the initial cluster center and the initial clustering group is accurate, so each initial clustering group is generated into multiple data groups. On the contrary, if there is any initial clustering group whose group cluster center is different from its initial cluster center, the group cluster center is updated to a new initial cluster center, and other data are used as new data to be clustered. Then, for the new initial cluster center and the new data to be clustered, the distance from each data to be clustered to each initial cluster center is calculated, and this cycle is repeated until the cluster center no longer changes, and the group cluster center of each initial cluster group is the same as its initial cluster center, and each initial cluster group is generated into multiple data groups.

[0124] This implementation performs a cyclic clustering analysis on the second target data and classifies the second target data with similar characteristics into the same group according to the characteristics of the second target data in production scale, production type, production process and production efficiency, thereby achieving accurate classification of the second target data.

[0125] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims. All equivalent structures or equivalent process changes made using the contents of the specification and drawings of the present invention, or directly or indirectly used in other related technical fields, are protected by the present invention.

Claims

1. A decision-making system based on big data optimization analysis, characterized in that: The decision-making system comprises: A data acquisition module, used to acquire first production process data of an enterprise, and crawl second production process data of a benchmark enterprise corresponding to the enterprise; A preprocessing module, used for performing denoising, deduplication and standardization processing on the first production process data and the second production process data respectively to obtain first target data and second target data; an analysis module, configured to perform cluster analysis on the second target data, divide the second target data into a plurality of data group classes, and determine a target data group class that matches the first target data in each of the data group classes; A decision suggestion module is used to determine the adjacent data group class corresponding to the target data group class, and generate decision suggestion information corresponding to the production process of the enterprise based on the adjacent data group class, so as to optimize the production process of the enterprise based on the decision suggestion information.

2. The decision-making system according to claim 1, characterized in that: The decision suggestion module comprises: A judging unit, configured to judge whether a first production efficiency corresponding to the adjacent data group class is higher than a second production efficiency corresponding to the target data group class; A search unit, configured to search for reference benchmarking enterprises from which each element in the adjacent data group class originates, and search for reference production process data corresponding to each of the reference benchmarking enterprises, when the first production efficiency is higher than the second production efficiency; A generating unit is used to generate the decision suggestion information based on the reference production process data and the first production process data.

3. The decision-making system according to claim 1, characterized in that: The analysis module comprises: A comparison unit, used for comparing the enterprise production characteristics between the second production process data corresponding to each of the data groups and the first production process data, to obtain a plurality of comparison results, wherein the enterprise production characteristics at least include production scale characteristics, production type characteristics, production process characteristics and production efficiency characteristics; The first determination unit is used to determine a target comparison result among the comparison results, and determine the data group class corresponding to the target comparison result as the target data group class, wherein the target comparison result is a comparison result of a successful match of enterprise production characteristics between the second production process data and the first production process data.

4. The decision-making system according to any one of claims 1 to 3, characterized in that: The analysis module also includes: a second determining unit, configured to randomly determine a plurality of second target data from the second target data as initial cluster centers, and use other data in the second target data except the initial cluster centers as data to be clustered; A calculation unit is used to calculate the distance between each of the to-be-clustered data and each of the initial cluster centers, and generate an initial cluster group according to the distance. The calculation formula of the distance is: Where d represents the distance, n represents the number of initial cluster centers, x represents the data to be clustered, and ci represents the i-th initial cluster center; The second determination unit is further used to determine the group cluster center of each of the initial cluster groups, and determine whether the group cluster center is the same as the initial cluster center, and if they are the same, generate each of the initial cluster groups into a plurality of the data group classes; An execution unit is used for, if the group cluster center is different from the initial cluster center, taking the group cluster center as a new initial cluster center, and executing the step of calculating the distance between each of the to-be-clustered data and each of the initial cluster centers.

5. The decision-making system according to any one of claims 1 to 3, characterized in that: The decision-making system also includes: A storage module is used to store the decision suggestion information, the adjacent data group class and the first production process data.

6. A decision-making method based on big data optimization analysis, characterized in that: The decision-making method includes: Acquire the first production process data of the enterprise, and crawl the second production process data of the benchmark enterprise corresponding to the enterprise; Performing denoising, deduplication and standardization processing on the first production process data and the second production process data respectively to obtain first target data and second target data; Performing cluster analysis on the second target data based on a preset analysis model, dividing the second target data into a plurality of data group classes, and determining a target data group class matching the first target data in each of the data group classes; Determine a neighboring data group class corresponding to the target data group class, and generate decision suggestion information corresponding to the production process of the enterprise based on the neighboring data group class, so as to optimize the production process of the enterprise based on the decision suggestion information.

7. The decision-making system according to claim 6, characterized in that: The step of generating decision suggestion information corresponding to the production process of the enterprise based on the adjacent data group class includes: Determining whether a first production efficiency corresponding to the adjacent data group class is higher than a second production efficiency corresponding to the target data group class; If the first production efficiency is higher than the second production efficiency, then searching for reference benchmarking enterprises from which each element in the adjacent data group class originates, and searching for reference production process data corresponding to each of the reference benchmarking enterprises; The decision suggestion information is generated based on the reference production process data and the first production process data.

8. The decision-making system according to claim 6, characterized in that: The step of determining a target data group class matching the first target data in each of the data group classes comprises: Comparing the enterprise production characteristics between the second production process data corresponding to each of the data groups and the first production process data to obtain a plurality of comparison results, wherein the enterprise production characteristics at least include production scale characteristics, production type characteristics, production process characteristics and production efficiency characteristics; Determine a target comparison result among the comparison results, and determine the data group class corresponding to the target comparison result as the target data group class, wherein the target comparison result is a comparison result in which the enterprise production characteristics between the second production process data and the first production process data are successfully matched.

9. The decision-making system according to any one of claims 6 to 8, characterized in that: The step of performing cluster analysis on the second target data based on a preset analysis model to divide the second target data into a plurality of data groups comprises: Randomly determine a plurality of second target data from the second target data as initial cluster centers, and use other data in the second target data except the initial cluster centers as data to be clustered; The distance between each of the data to be clustered and each of the initial cluster centers is calculated, and an initial cluster group is generated according to the distance. The calculation formula of the distance is: Where d represents the distance, n represents the number of initial cluster centers, x represents the data to be clustered, and ci represents the i-th initial cluster center; Determine the group cluster center of each of the initial cluster groups, and determine whether the group cluster center is the same as the initial cluster center, and if they are the same, generate each of the initial cluster groups into a plurality of the data group classes; If the group cluster center is different from the initial cluster center, the group cluster center is used as a new initial cluster center, and the step of calculating the distance from each of the to-be-clustered data to each of the initial cluster centers is performed.

10. The decision-making system according to any one of claims 6 to 8, characterized in that: After the step of generating decision suggestion information corresponding to the production process of the enterprise based on the adjacent data group class, the following step is further included: The decision suggestion information, the neighboring data group class and the first production process data are stored.

Images