User demand data processing method and device

Abstract

The embodiment of the present invention provides a method and device for processing user demand data, wherein the method comprises: obtaining demand data of a target object; performing standardization processing on the demand data to obtain standardized demand data; performing data calibration on the standardized demand data to obtain calibration data; performing cluster calculation on the calibration data to obtain cluster calculation results. The present invention solves the problem that the service fee calculation method is not accurate enough, and achieves the effect of improving the accuracy of the service fee calculation method.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of data processing, and in particular, to a method and device for processing user demand data. Background Art

[0002] Currently, the service billing of most service providers is based either on cost-value or market competition. However, there is a big difference between innovative services and physical services. Many innovative services are online services, data services, and resource services, and their costs are difficult to determine. Current market pricing also lacks theoretical explanations and data support, and is not reasonable or scientific.

[0003] Currently, most patents about "billing models" are about traffic billing, electricity price billing or the design and development of billing systems, such as: "A capability open metering and billing model for multiple users and multiple scenarios", "A real-time billing management method and system for cloud resources" and other solutions.

[0004] Regarding "service billing", there is a patent for "method and device for calculating service fees". This patent provides a method and device for calculating service fees, wherein a method for calculating service fees includes: obtaining a collection event of a sub-program added to a service project; the collection event contains a program identifier, a service category and fee information; based on the program identifier contained in the collection event, determining whether the sub-program has signed a service agreement; if so, constructing a service billing event based on the collection event, and matching the service billing event with the service agreement to determine the service fee rate of the service category; calculating the service fee of the service billing event based on the fee information and the service fee rate.

[0005] It can be seen from this that due to the inability to determine the content of user needs, there is a problem in the relevant technology that the method of calculating service fees is not accurate enough.

[0006] Currently, no effective solution has been proposed for the above-mentioned problems existing in the related technologies. Summary of the invention

[0007] The embodiment of the present invention provides a method and device for processing user demand data, so as to at least solve the problem that the calculation method of service fee in the related art is not accurate enough.

[0008] According to one embodiment of the present invention, a method for processing user demand data is provided, comprising: acquiring demand data of a target object; performing standardization processing on the demand data to obtain standardized demand data; performing data calibration on the standardized demand data to obtain calibration data; performing cluster calculation on the calibration data to obtain cluster calculation results.

[0009] Furthermore, after clustering the calibration data to obtain clustering calculation results, the method further includes: determining charging parameters for the target object based on the clustering calculation results, wherein the charging parameters are used to identify fees to be paid by the target object to the service provider.

[0010] Furthermore, after performing clustering calculation on the calibration data and obtaining the clustering calculation results, the method also includes: obtaining parameters for clustering calculation, wherein the parameters include at least the number of classifications c, the exponential weight r and the initial classification fuzzy matrix A(0); performing clustering calculation on the calibration data according to the parameters to obtain the clustering results.

[0011] Furthermore, after performing cluster calculation on the calibration data and obtaining the cluster calculation results, the method also includes: evaluating the cluster calculation results according to preset evaluation indicators to obtain evaluation results; wherein the preset evaluation indicators include a partition coefficient, an average fuzzy entropy of the partition, and a proportional index; and correcting the model parameters of the cluster calculation according to the evaluation results.

[0012] Furthermore, performing data calibration on the demand standardized data to obtain the calibration data includes: acquiring a pre-established fuzzy similarity matrix; and performing data calibration on the demand standardized data through the fuzzy similarity matrix to obtain the calibration data.

[0013] Furthermore, before obtaining the pre-established fuzzy similarity matrix, the method also includes: obtaining a predetermined number of sample data, wherein each sample data contains demand data of the target object; determining the similarity matrix of the clustering algorithm based on the sample data; and determining the pre-established fuzzy similarity matrix through the similarity matrix of the clustering algorithm.

[0014] Furthermore, performing clustering calculation on the calibration data to obtain the clustering calculation result includes: performing clustering calculation on the calibration data by any one of a transitive closure method, a maximum tree method, and a dynamic direct clustering method to obtain the clustering calculation result.

[0015] According to another embodiment of the present invention, a user demand data processing device is provided, including: an acquisition unit, used to acquire the demand data of a target object; a standard unit, used to perform standardization processing on the demand data to obtain demand standardized data; a calibration unit, used to perform data calibration on the demand standardized data to obtain calibration data; and a calculation unit, used to perform cluster calculation on the calibration data to obtain a cluster calculation result.

[0016] According to yet another embodiment of the present invention, a computer-readable storage medium is provided, in which a computer program is stored, wherein the computer program is configured to execute the steps of any one of the above method embodiments when run.

[0017] According to yet another embodiment of the present invention, there is provided an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0018] The present invention obtains demand data of the target object; standardizes the demand data to obtain standardized demand data; calibrates the standardized demand data to obtain calibration data; and performs cluster calculation on the calibration data to obtain cluster calculation results. Therefore, accurate cluster calculation results can be obtained based on the demand data. Since the calculation results are accurate, the problem of inaccurate service fee calculation methods in related technologies can be solved, thereby achieving the effect of improving the accuracy of the service fee calculation method. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a hardware structure block diagram of a mobile terminal of a method for processing user demand data according to an embodiment of the present invention;

[0020] Figure 2 is a flowchart of a method for processing user demand data according to an embodiment of the present invention;

[0021] Figure 3 is a schematic diagram of an original data matrix of user requirements according to an embodiment of the present invention;

[0022] Figure 4 is another schematic diagram of an original data matrix of user requirements according to an embodiment of the present invention;

[0023] Figure 5 4 is a structural block diagram of a user demand data processing device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0024] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings and in combination with the embodiments.

[0025] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0026] The method embodiments provided in the embodiments of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Taking running on a mobile terminal as an example, Figure 1FIG. 1 is a hardware structure block diagram of a mobile terminal of a method for processing user demand data according to an embodiment of the present invention. Figure 1 As shown, the mobile terminal may include one or more ( Figure 1 Only one is shown in the figure) a processor 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data, wherein the mobile terminal may also include a transmission device 106 and an input / output device 108 for communication functions. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the mobile terminal. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations shown.

[0027] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the user demand data processing method in the embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, to implement the above method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the mobile terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0028] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the mobile terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0029] In this embodiment, a method for processing user demand data is provided. Figure 2 is a flow chart of a method for processing user demand data according to an embodiment of the present invention. Figure 2 As shown, the process includes the following steps:

[0030] Step S101, obtaining the demand data of the target object;

[0031] Step S102, standardizing the demand data to obtain standardized demand data;

[0032] Step S103, calibrating the required standardized data to obtain calibrated data;

[0033] Step S104, performing cluster calculation on the calibration data to obtain a cluster calculation result.

[0034] In the above embodiment, the demand data of the target object is obtained; the demand data is standardized to obtain the demand standardized data; the demand standardized data is calibrated to obtain the calibration data; the calibration data is clustered to obtain the cluster calculation result. Through the present invention, the problem of inaccurate calculation method of service fee is solved, and the accuracy of the calculation method of service fee is improved.

[0035] Optionally, after clustering calculation is performed on the calibration data to obtain a clustering calculation result, a charging parameter for the target object is determined based on the clustering calculation result, wherein the charging parameter is used to identify a fee to be paid by the target object to the service provider.

[0036] In this embodiment, the target object can be an individual user or demand data of an enterprise, etc. The demand data is demand data for software services or other types of services. The demand data is standardized to obtain standardized data that conforms to software development services. This can improve data processing efficiency and reduce manual processing time and costs. The standardized demand data is calibrated, and then the calibrated data is clustered to obtain a clustering calculation result. This clustering calculation result can be used to calculate service fees. Since the clustering calculation result can more accurately reflect the workload of the development process or the amount of service data, it can be used to calculate the charging parameters for the target object, and the cost of developing or using the meta-service data package is paid to the service provider based on the charging parameters, thereby improving the accuracy of the service fee calculation method.

[0037] Optionally, after performing clustering calculation on the calibration data and obtaining the clustering calculation results, the parameters of the clustering calculation are obtained, wherein the parameters include at least the number of classifications c, the exponential weight r and the initial classification fuzzy matrix A(0); clustering calculation is performed on the calibration data according to the parameters to obtain the clustering results.

[0038] Clustering calculation can use ISODATA fuzzy clustering analysis method, which requires the number of classifications c, index weight r and initial classification fuzzy matrix A(0) to be determined in advance. The final classification is closely related to the values ​​of the above parameters. Clustering calculation can be performed on the calibration data according to the parameters to obtain the clustering results.

[0039] Optionally, after performing cluster calculation on the calibration data and obtaining the cluster calculation results, the cluster calculation results are evaluated according to preset evaluation indicators to obtain evaluation results; wherein the preset evaluation indicators include the partition coefficient, the average fuzzy entropy of the partition and the proportional index; and the model parameters of the cluster calculation are corrected according to the evaluation results.

[0040] In order to improve the accuracy of clustering calculation results, the clustering calculation results can be evaluated. The evaluation is performed based on various evaluation parameters. The model or model parameters used in the clustering calculation are corrected through the evaluation results, which can improve the accuracy of the clustering calculation results.

[0041] Optionally, performing data calibration on the demand standardized data to obtain the calibration data includes: acquiring a pre-established fuzzy similarity matrix; and performing data calibration on the demand standardized data through the fuzzy similarity matrix to obtain the calibration data.

[0042] Optionally, before obtaining the pre-established fuzzy similarity matrix, a predetermined number of sample data are obtained, wherein each sample data contains the demand data of the target object; a similarity matrix of the clustering algorithm is determined based on the sample data; and the pre-established fuzzy similarity matrix is ​​determined through the similarity matrix of the clustering algorithm.

[0043] The calibration data can be calculated through a pre-established fuzzy similarity matrix. The fuzzy similarity matrix is ​​obtained by training multiple groups of sample data, which can be obtained through training of a convolutional neural network model. Calibration of the required standardized data through the trained fuzzy similarity matrix can improve calculation efficiency and save manual calculation costs.

[0044] Optionally, performing clustering calculation on the calibration data to obtain the clustering calculation result includes: performing clustering calculation on the calibration data by any one of a transitive closure method, a maximum tree method, and a dynamic direct clustering method to obtain the clustering calculation result.

[0045] Cluster computing can be implemented through a variety of methods, which can be selected according to the specific usage scenario.

[0046] This embodiment can be used as a billing system for an innovation and entrepreneurship service platform based on meta-services. In the above embodiment, demand information refers to the demand information of users or demanders. The demand information usually contains the operations to be performed, the functions to be implemented, etc. Some meta-service packages can be recombined according to the demand information to obtain an integrated service package that meets the demand information. The integrated service package can be used in specific scenarios after some configuration. The configuration is performed according to the specific scenario to be used by the specific user. Some contents of the integrated service package are configured according to the demand information to obtain a target service package that meets the user's needs. Finally, the service fee parameter is calculated according to the number of meta-service packages in this service package and the level of each meta-service package. This parameter can be directly the amount of a certain currency or other units of measurement. Among them, the level of the meta-service package is pre-set by the service provider to distinguish the workload of different meta-service packages in order to calculate the fee.

[0047] It should be noted that, in this embodiment, through such a billing method, the cost result can be obtained comprehensively based on the content and level of the service package provided by the service provider, so that the determination of the cost is more reasonable.

[0048] This embodiment also provides a specific implementation method, which will be discussed below in conjunction with the specific implementation method.

[0049] Application of fuzzy clustering analysis method in demand identification:

[0050] 1. Implementation technology of fuzzy clustering analysis in user demand identification.

[0051] 1.1 Implementation technology of fuzzy equivalence matrix

[0052] (1) General steps in identifying user needs:

[0053] Step 1: Clarify the analysis object and prepare the original data.

[0054] Fuzzy cluster analysis is an application technology that ultimately serves the identification of user needs. Therefore, the object of cluster analysis should be proposed based on the user's needs, which is generally extracted from the user's requirements for product attributes. Fuzzy cluster analysis has certain requirements for data, such as low data redundancy, low correlation between data attributes, and low data error rate. However, the data collected in the real world usually has the characteristics of chaos, redundancy, and incompleteness. Therefore, cluster analysis must go through the data preparation stage to improve the quality of analysis.

[0055] Step 2: Standardization of user demand data.

[0056] (1) User demand data matrix Let the user demand set X = {x1, x2, …, x n}, each requirement is composed of m user requirement characteristic indicators, namely: xi = {x i1 ,x i1 ,…,x in}(i=1,2,…,n), thus, we get the original data matrix of user requirements, such as Figure 3 shown.

[0057] (2) User demand data standardization:

[0058] In the actual user demand set, different demand data generally have different dimensions. In order to make the quantities with different dimensions comparable, it is usually necessary to make appropriate transformations to the demand data. However, even so, the data obtained may not be in the interval [0,1]. Therefore, the standardization of user demand data mentioned here is to compress the data to the interval [0,1] according to the requirements of the fuzzy matrix. Usually the following two transformations are required:

[0059] ① Translation and standard deviation transformation

[0060]

[0061] in,

[0062] After the transformation, the mean of each variable is 0, the standard deviation is 1, and the dimension effect is eliminated. However, the x obtained in this way is ki It is not necessarily in the interval [0,1].

[0063] ②Shift and range transformation

[0064]

[0065] Obviously It also eliminates the effect of dimension.

[0066] Step 3: Calibration (establishing fuzzy similarity matrix).

[0067] Assume that the user demand set X = {x1, x1,…, x n}, x i ={x i1 ,x i1 ,…,x in According to the traditional clustering method, the similarity matrix is ​​determined, and the demand xi and demand x j The similarity r ij =R(x i , x j ). Depending on the nature of the problem, one of the following formulas can be selected to calculate r ij .

[0068] (1) Angle cosine method

[0069]

[0070] (2) Correlation coefficient method

[0071]

[0072] in,

[0073] (3) Euclidean distance method

[0074]

[0075] (4) Subjective scoring method

[0076] Please have practical experience directly to the requirements x i With demand x j The similarity score of ij Of course, there are other methods, such as scalar product method, exponential similarity coefficient method, maximum minimum method, geometric mean minimum method, absolute value subtraction method, arithmetic mean minimum method, absolute value reciprocal method, absolute value exponential method, Hamming distance, Chebyshev distance, etc., which can be selected according to the nature of the problem and convenience of use.

[0077] Step 4: Clustering (finding a dynamic clustering diagram)

[0078] (1) Transitive Closure Method

[0079] The fuzzy matrix obtained by calibration is only a fuzzy similarity matrix R, which is not necessarily transitive, that is, R is not necessarily a fuzzy equivalence matrix. In order to classify, R needs to be transformed into a fuzzy equivalence matrix R*, and the transitive closure t(R) of R is obtained. This is the fuzzy equivalence matrix R*, that is, t(R) = R*. Then let λ change from large to small, and a dynamic clustering graph can be formed. The following method can be used to transform R into a fuzzy equivalence matrix R*:

[0080] ① Flat method

[0081] Since R is the fuzzy similarity matrix, R k is the fuzzy similarity matrix; according to the transitive closure definition, t(R) = R k is a transfer matrix; therefore, Rk is a fuzzy equivalent matrix, that is, the transfer closure t(R) of the fuzzy similarity matrix R is a fuzzy equivalent matrix. Starting from a fuzzy similarity matrix, a fuzzy equivalent matrix can be constructed by using the method of finding the transfer closure. In order to save the amount of calculation, the square method can be used:

[0082] R→R 2 →R 4 →R 2k ........

[0083] After a finite number of operations, there must be R 2k =R 2k+1 . So we get a fuzzy equivalent matrix t(R)=R 2k .

[0084] ② Direct solution method

[0085] Step 1: Find the representative element of R;

[0086] Step 2: Select the same fruit elements as some representative elements of R;

[0087] Step 3: Write down where t(R) = [r ij * ]

[0088] Among them, r ij * =r ij , when r ij is the representative element of R;

[0089] r ij * =r kl , when r ij It is not a representative element of R.

[0090] Here r k1 is the non-representative element r of R ij The standard element.

[0091] (2) Maximum Tree Method

[0092] The so-called maximum tree method is to build the fuzzy similarity matrix R of user needs, and then use a variety of methods to draw the maximum tree (i.e., the maximum support tree) of the weighted graph corresponding to R, and then perform clustering. Commonly used methods include: Prim method and Kruskal method. The maximum tree drawn by the above method may not be unique. But it can be proved that the classification results are the same.

[0093] (3) Dynamic direct clustering method

[0094] When the number of user demand sets increases slightly, the transitive closure method becomes quite troublesome. Dynamic direct clustering is a relatively simple method. First, the fuzzy similarity matrix R is established, then the primitives of R are obtained, and finally the dynamic clustering diagram is drawn.

[0095] The algorithm for finding R primitives is as follows:

[0096] The first step is to mark the double pivots of R (n-order fuzzy similarity matrix) with double boxes, mark the single pivots of R with single boxes, and write out the row-label decomposition set of R:

[0097] H1,H2,…Hr.

[0098] Step 2: If i,j∈Hk(k=1,…P), and rij is not the main element of R (called a connected element), mark it with ﹏.

[0099] Step 3: Select the largest element among the unmarked elements of R (if there are several maximal elements, select any one rkl from them, called the connecting element of R, and mark it with -. If p-1 connecting elements have been selected, the algorithm stops; otherwise, go to step 4.

[0100] Step 4: Use the connection element rkl to merge the existing code set. Also, clean up R and then go to step 3.

[0101] 2. Determination of the optimal threshold value λ

[0102] In fuzzy cluster analysis, for different λ∈[0,1], different user classifications can be obtained, thus forming a dynamic clustering diagram, which is more vivid and intuitive for understanding the classification of user demand attributes. However, many user demand problems require the selection of a certain domain value λ to determine a specific classification of the sample. This raises the question of how to determine the domain value λ. The following are two methods:

[0103] (1) According to actual needs, in the dynamic clustering diagram, the value of λ is adjusted to obtain appropriate classification, without the need to accurately estimate in advance how many categories the samples should be classified into. Of course, experienced experts can also combine professional knowledge to determine the domain value λ, thereby deriving an equivalent relationship at the λ level.

[0104] (2) Use F-statistics to determine the optimal value of λ. Assume that the user demand set U = {x1, x2, ... xn}, and each specific demand xi has m user demand characteristic indicators: xi = {xi1, xi2, ... xim} (i-1, 2, ..., n). Thus, the original user demand data matrix is ​​obtained, as Figure 4 shown.

[0105] in, It is called the central vector of the overall user demand set.

[0106] Suppose the number of categories corresponding to the value is r, the number of samples in the jth category is nj, and the samples in the jth category are recorded as: x1 (j) ,x2 (j) ,x nj (j) , the cluster center of the jth class is the vector: in, is the average value of the kth user demand feature:

[0107] To do the F-statistic:

[0108]

[0109] in, for and The distance is the jth class sample With Center distance.

[0110] Its numerator represents the distance between the features and the class, and the denominator represents the distance between class samples. Therefore, the larger the F value, the more reasonable the classification, and the domain value λ corresponding to the maximum F-statistic is the optimal value.

[0111] 1.2 Implementation Technology of Fuzzy Partition

[0112] 1. User needs are diverse, and the importance of each factor is fuzzy. In order to determine the importance of each demand factor in user needs, the fuzzy c-partition method can be used to reasonably divide the space of user demand factors.

[0113] Assume that the user demand set X = {x1, x2, ..., xn} is a finite set, which is the sample for cluster analysis. Divide the sample into c fuzzy subsets, and obtain the fuzzy c-partition space of X:

[0114]

[0115] It satisfies:

[0116] (1)

[0117] (2)

[0118] (3)

[0119] Among them, c is the number of categories of user demand factors, For xj belongs to the class degree of affiliation.

[0120] So the fuzzy c-partition of x can be represented by a c×n fuzzy matrix A=(a ij ) c×n Indicates that Then, each fuzzy matrix A corresponds to a fuzzy c-partition of X, and all fuzzy c-partitions are called the partition space of X, denoted by Mjc.

[0121] 2 Determination of cluster centers and cluster targets

[0122] The direct purpose of clustering the user requirement element X is to select the "optimal" classification fuzzy matrix from the partition space Mjc of X. To this end, the clustering centers of each classification should be obtained. The clustering center is the "core" of all elements in each classification. In a reasonable classification fuzzy matrix, the sum of the squares of the distances from each element in a class to the clustering center of that class should be as small as possible.

[0123] Suppose there are p features for the elements in the user requirement element X, so xj can be expressed as:

[0124] x j =(x j1 ,x j2 ,…x jp ), j = 1, 2, … n

[0125] Also suppose the c clustering centers of the user requirement element X are V = {v1, v2, … v ip}, i = 1, 2, … c

[0126] The clustering center can be calculated according to the following formula:

[0127]

[0128] In the formula, r is a to-be-determined parameter representing the exponential weight. Generally, r > 1. The larger r is, the more fuzzy the classification is. When r → 1, the fuzziness of the classification is smaller and finally becomes a hard classification, that is, the partition of X becomes a clear c - partition.

[0129] The "optimal" fuzzy c - partition of the user requirement element can be obtained by minimizing the following objective function:

[0130] 3 Clustering process

[0131] The above formula is actually a non - linear programming problem. It is relatively difficult to directly find its optimal solution. In practical problems, an initial classification fuzzy matrix will be given in advance, and it will be continuously improved using a certain iterative rule until the iterative termination criterion is met. The specific iterative process is as follows.

[0132] Step1: Select c (2 ≤ c ≤ n), r (1 < r < +∞), set the iteration times variable l = 0, and randomly determine the initial classification fuzzy matrix A(0) ∈ Mjc.

[0133] Step2: Calculate the clustering center v i (l) (i = 1, 2, … c)

[0134] Step3: When x j ≠v i (l) (i = 1, 2, … c), according to vi (l) (i=1,2,…c) calculate a ij (l+1)

[0135]

[0136] Otherwise, let a ij (l+1) =1(if j=i), a ij (l+1) =0 (if j≠i), then we get the new fuzzy matrix A (l+1) =(a ij (l+1) ) c×n ;

[0137] Step 4: Compare A (l+1) and A (l) , for a given infinitesimal positive number ε>0, if the matrix norm ||A (l+1) -A (l) ||>ε, set l=l+1 and go to step 2; otherwise, stop the iteration.

[0138] So far, we have obtained a relatively ideal fuzzy c-partition matrix, recorded as A*. Next, we will use A* to finally determine which category each element in the user demand factor X belongs to. The classification method is like Then xi is classified into the i0th class, where Vio is the cluster center of the i0th class.

[0139] 4 Clustering effect evaluation

[0140] Since the ISODATA fuzzy clustering analysis method requires the pre-determination of the number of classifications c, the index weight r and the initial classification fuzzy matrix A(0), the final classification is closely related to the values ​​of the above data. Therefore, the division of the user demand factor X obtained is not necessarily optimal. It is necessary to evaluate the final results of the clustering analysis. Three indicators, such as the division coefficient, the average fuzzy entropy of the division and the proportional index, can be considered to evaluate the clustering effect of the fuzzy c-division.

[0141] (1) Division coefficient F(A,c)

[0142] The quantitative description of the uncertainty degree of fuzzy division is calculated as follows:

[0143] And there is The closer the partition coefficient is to 1, the smaller the fuzziness of the classification is, which means the better the clustering effect is.

[0144] (2) Average fuzzy entropy H(A,c) of the partitioned area

[0145]

[0146] 0≤H(A,c)≤lnc, the smaller the average fuzzy entropy, the smaller the uncertainty of the final classification, and the better the clustering effect.

[0147] (3) Proportional index P(A,c)

[0148]

[0149] Among them, α j =max 1≤i≤c {a ij},[a j -1 ] is not greater than a j -1 The largest integer,

[0150] 0≤P(A,C)<+∞, the larger the proportional index, the better the fuzzy clustering effect.

[0151] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0152] In this embodiment, a user demand data processing device is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware of a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0153] Figure 5 is a structural block diagram of a user demand data processing device according to an embodiment of the present invention. Figure 5 As shown, the device comprises:

[0154] An acquisition unit 10 is used to acquire the demand data of the target object;

[0155] The standard unit 20 is used to perform standardization processing on the demand data to obtain demand standardized data;

[0156] The calibration unit 30 is used to calibrate the required standardized data to obtain calibration data;

[0157] The calculation unit 40 is used to perform cluster calculation on the calibration data to obtain a cluster calculation result.

[0158] The technical solution of this embodiment can solve the problem of inaccurate service fee calculation method in related technologies, thereby achieving the effect of improving the accuracy of the service fee calculation method.

[0159] It should be noted that the above modules can be implemented by software or hardware. For the latter, it can be implemented in the following ways, but not limited to: the above modules are all located in the same processor; or the above modules are located in different processors in any combination.

[0160] An embodiment of the present invention further provides a computer-readable storage medium, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above method embodiments when running.

[0161] In an exemplary embodiment, the computer-readable storage medium may include, but is not limited to, various media that can store computer programs, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk.

[0162] An embodiment of the present invention further provides an electronic device, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0163] In an exemplary embodiment, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0164] For specific examples in this embodiment, reference may be made to the examples described in the above embodiments and exemplary implementation modes, and this embodiment will not be described in detail herein.

[0165] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0166] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for processing user demand data, characterized in that: include: Obtain the target object's demand data; Performing standardization processing on the demand data to obtain demand standardized data; Performing data calibration on the required standardized data to obtain calibration data; Performing cluster calculation on the calibration data to obtain a cluster calculation result; Determining charging parameters for the target object based on the cluster calculation result, wherein the charging parameters are used to identify fees to be paid by the target object to the service provider; Obtaining parameters for clustering calculation, wherein the parameters include at least the number of classifications c, the exponential weight r, and the initial classification fuzzy matrix A(0); Performing cluster calculation on the calibration data according to the parameters to obtain the cluster calculation result; The clustering calculation result is evaluated according to preset evaluation indicators to obtain an evaluation result; wherein the preset evaluation indicators include a partition coefficient, an average fuzzy entropy of the partition, and a proportional index; Modifying the model parameters of the clustering calculation according to the evaluation results; The required standardized data is calibrated to obtain the calibrated data including: Obtain a pre-established fuzzy similarity matrix; Performing data calibration on the demand standardized data by using the fuzzy similarity matrix to obtain the calibration data; Before obtaining the pre-established fuzzy similarity matrix, the method further comprises: Acquire a predetermined number of sample data, wherein each sample data contains the demand data of the target object; Determining a similarity matrix of a clustering algorithm based on the sample data; The pre-established fuzzy similarity matrix is ​​determined by the similarity matrix of the clustering algorithm.

2. The method according to claim 1, characterized in that Performing cluster calculation on the calibration data to obtain cluster calculation results including: The calibration data is clustered by any one of a transitive closure method, a maximum tree method, and a dynamic direct clustering method to obtain a clustering calculation result.

3. A user demand data processing device, characterized in that: include: An acquisition unit, used for acquiring demand data of a target object; A standard unit, used for performing standardization processing on the demand data to obtain demand standardized data; The calibration unit is used to calibrate the demand standardized data to obtain the calibration data, including: obtaining a pre-established fuzzy similarity matrix; calibrating the demand standardized data through the fuzzy similarity matrix to obtain the calibration data; before obtaining the pre-established fuzzy similarity matrix, it also includes: obtaining a predetermined number of sample data, wherein each sample data contains the demand data of the target object; determining the similarity matrix of the clustering algorithm according to the sample data; determining the pre-established fuzzy similarity matrix through the similarity matrix of the clustering algorithm; A calculation unit is used to perform cluster calculation on the calibration data to obtain cluster calculation results, determine charging parameters for the target object based on the cluster calculation results, wherein the charging parameters are used to identify the fees to be paid by the target object to the service provider; obtain parameters for cluster calculation, wherein the parameters at least include the number of classifications c, the exponential weight r and the initial classification fuzzy matrix A(0); perform cluster calculation on the calibration data according to the parameters to obtain the cluster calculation results; evaluate the cluster calculation results according to preset evaluation indicators to obtain evaluation results; wherein the preset evaluation indicators include a partition coefficient, an average fuzzy entropy of the partition and a proportional index; and modify the model parameters of the cluster calculation according to the evaluation results.

4. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program is configured to execute the method according to any one of claims 1 to 2 when executed.

5. An electronic device comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to run the computer program to perform the method according to any one of claims 1 to 2.

Images