A demand-driven product evaluation index system optimization method and model

Abstract

The present invention belongs to the technical field of product evaluation, and specifically relates to a demand-driven product evaluation index system optimization method and model. The demand-driven product evaluation index system optimization method includes: determining the final index to be optimized; optimizing the final index based on the clustering relationship, correlation coefficient and constraint control variable value between the final index to be optimized; performing cluster analysis on the optimized lower-level index; performing affiliation analysis on the optimized lower-level index and the original upper-level index; optimizing the upper-level index based on the optimization control factor of the lower-level index of the same cluster in the optimized lower-level index and the proportion of the newly added index; optimizing the upper-level index step by step until all the indicators are optimized. The present invention can optimize the final index and the upper-level index, is suitable for the optimization of a multi-level index system, covers three different types of index deletion, addition and replacement, takes into account the influence of multiple factors during optimization, and the optimization path is more scientific, reasonable and comprehensive.

Description

Technical Field

[0001] The present invention belongs to the technical field of product evaluation, and specifically relates to a demand-driven product evaluation index system optimization method and model. Background Art

[0002] At present, industries such as automobiles, construction machinery, and mobile phones have established or are in the process of establishing product evaluation systems. The core of the product evaluation system is the evaluation index system. After the product evaluation index system is established and operated for a period of time, it is often affected by product user needs, market regulations, etc., resulting in the original evaluation indicators being unable to truly reflect user demands. The error between the evaluation results and market feedback gradually increases, resulting in distorted results. At this time, the evaluation index system needs to be improved and optimized.

[0003] At present, the index screening is mainly carried out through index weight sorting, cluster analysis, correlation analysis, sensitivity significance analysis, rough set theory and other methods, and then the selected indexes are merged to form an optimized index system. For related technical solutions, please refer to: CN109544026A Optimization of consistency index of waste power batteries based on clustering and entropy ratio, CN111080108B Data-driven weapon equipment combat effectiveness evaluation index screening method and system, CN104732447B A method for establishing a vulnerability index system for important power grid infrastructure, CN201510802537.4 A method and system for optimizing a company's operating index system, CN201810049149.7 A risk assessment method for water and mud inrush in underground engineering based on multivariate precursor information. The existing index screening methods mainly adopt the method of deleting duplicate and redundant indicators to screen indicators, while the mature multi-level index system that has been built rarely has the problem of duplicate and redundant indicators. When the current index screening method is applied to the screening of new indicators and original indicators, it cannot directly reflect the user demand factors that affect the final determination of whether the indicator is retained, the technical personnel's product development demand factors, etc. That is, the update of the existing index system cannot fully reflect the changes in relevant influencing factors, and the addition, deletion and replacement of indicators have great uncertainty, which greatly affects the effectiveness of the final product evaluation results. In addition, the existing index system optimization method cannot effectively optimize the upper-level indicators.

[0004] Therefore, it is urgent to propose a scientific and reasonable method for screening and optimizing product evaluation indicators. Summary of the invention

[0005] In order to solve the shortcomings of the prior art, the present invention provides a demand-driven product evaluation index system optimization method, which includes the optimization of the final index and the upper index, is suitable for the optimization of a multi-level index system, and is more comprehensive and reasonable.

[0006] Another object of the present invention is to provide a demand-driven product evaluation index system optimization model, which can optimize the final index and the upper index, is suitable for the optimization of a multi-level index system, and is more comprehensive and reasonable.

[0007] In order to solve the deficiencies of the prior art, the technical solution provided by the present invention is:

[0008] A demand-driven product evaluation index system optimization method, comprising:

[0009] S1: Determine the final level indicator to be optimized; perform cluster analysis and correlation analysis on the final level indicator to be optimized, and calculate the constraint control variable value of the final level indicator to be optimized; optimize the final level indicator based on the cluster relationship, correlation coefficient and constraint control variable value between the final level indicators to be optimized;

[0010] S2: Perform cluster analysis on the optimized subordinate indicators; perform affiliation analysis on the optimized subordinate indicators and the original superior indicators; optimize the superior indicators based on the optimized control factors of the subordinate indicators of the same cluster and the proportion of the newly added indicators in the optimized subordinate indicators;

[0011] S3: Repeat step S2 to optimize the upper-level indicators step by step until all indicators are optimized.

[0012] Preferably, the determining of the final level indicator to be optimized includes:

[0013] Determine new terminal indicators based on user needs, quality feedback, R&D needs, and standards and regulations;

[0014] The original terminal indicators and newly added terminal indicators are identified as terminal indicators to be optimized.

[0015] Preferably, the optimization of the final level indicators based on the clustering relationship, correlation coefficient and constraint control variable value between the final level indicators to be optimized includes:

[0016] Determine whether the number of final-level indicators to be optimized in the same cluster is greater than 1. If so, optimize the final-level indicators to be optimized in the cluster based on the correlation coefficients between the final-level indicators to be optimized and the values ​​of the constraint control variables; otherwise, optimize the final-level indicators to be optimized in the cluster based on the values ​​of the constraint control variables of the final-level indicators to be optimized.

[0017] Preferably, the optimizing the final level index to be optimized in the cluster based on the correlation coefficient between the final level indexes to be optimized and the constraint control variable value includes:

[0018] judge Is it greater than or equal to r0? If so, delete x i 、x j The final indicator to be optimized has a small value of the medium constraint control variable;

[0019] Otherwise, if exist Arrange the first n1 digits in descending order, then keep x i , otherwise delete x i ;

[0020] in,

[0021] x i 、x j are the i-th and j-th final-level indicators to be optimized, respectively, i, j = 1, ..., n; n is the total number of final-level indicators to be optimized;

[0022] For x i With x j The correlation coefficient; r0 is the strong correlation determination threshold;

[0023] For x i With x j The ratio of the values ​​of the constraint control variables; For x i The sum of the ratios of the constraint control variable values ​​of each final level indicator to be optimized in the cluster; m s is the number of final-level indicators to be optimized in the s-th cluster; is the final indicator to be optimized in the same cluster The matrix composed of The calculation formula is:

[0024]

[0025] in, x i 、x j The constraint control variable value is calculated as follows:

[0026]

[0027]

[0028] x i 、x j The relevance of user needs;

[0029] x i 、x j The relevance of technology requirements;

[0030] x i 、x j correction factor.

[0031] Preferably, the optimizing the final level index to be optimized of the cluster based on the constraint control variable value between the final level indexes to be optimized includes:

[0032] like exist If the descending order is the first n2, then keep x i , otherwise delete x i ;

[0033] Among them, x i 、x j are the i-th and j-th final-level indicators to be optimized, respectively, i, j = 1, ..., n; n is the total number of final-level indicators to be optimized;

[0034] For x i With x j The ratio of the values ​​of the constraint control variables; For x i The sum of the ratios of the constraint control variable values ​​of each final-level indicator to be optimized; The final indicator to be optimized The matrix composed of The calculation formula is:

[0035]

[0036] in, x i 、x j The constraint control variable value is calculated as follows:

[0037]

[0038]

[0039] x i 、x j The relevance of user needs;

[0040] x i 、x j The relevance of technology requirements;

[0041] x i 、x j correction factor.

[0042] Preferably, the cluster analysis of the optimized lower-level indicators includes: dividing the optimized lower-level indicators into n x clusters, where the ath cluster is represented by X a; Make new indicator judgments on the subordinate indicators in each cluster and determine the new indicators in each cluster;

[0043] The subordination analysis of the optimized lower-level indicators and the original upper-level indicators includes: determining the original upper-level indicators to which each cluster belongs for the other clusters except for the clusters that all contain the newly added indicators;

[0044] The stepwise optimization of the upper level indicators based on the optimization control factors of the lower level indicators of the same cluster and the proportion of the newly added indicators in the optimized lower level indicators includes:

[0045] Determine the ath cluster of the optimized subordinate indicators Is it greater than δ0? If so, define a new parent indicator for the cluster;

[0046] Otherwise, determine the a-th cluster Is it greater than p0? If so, define a new parent indicator for the cluster;

[0047] Otherwise, the original superior indicator to which the cluster belongs is used as the superior indicator of the cluster;

[0048] in, is the optimization control factor of the ath cluster of the optimized subordinate indicators; δ0 is the optimization control judgment threshold; is the proportion of newly added indicators in the ath cluster of the optimized subordinate indicators; p0 is the judgment threshold of the newly added indicators.

[0049] A demand-driven product evaluation index system optimization model, including a terminal index optimization module and a superior index optimization module;

[0050] The final index optimization module is used to determine the final index to be optimized; perform cluster analysis and correlation analysis on the final index to be optimized, and calculate the constraint control variable value of the final index to be optimized; optimize the final index based on the cluster relationship, correlation coefficient and constraint control variable value between the final indexes to be optimized;

[0051] The superior indicator optimization module is used to perform cluster analysis on the optimized subordinate indicators; perform affiliation analysis on the optimized subordinate indicators and the original superior indicators; optimize the superior indicators based on the optimization control factors of the subordinate indicators of the same cluster in the optimized subordinate indicators and the proportion of new indicators; optimize the superior indicators step by step until all indicators are optimized.

[0052] Preferably, the final-level indicator optimization module is used to determine whether the number of final-level indicators to be optimized in the same cluster is greater than 1. If so, the final-level indicators to be optimized in the cluster are optimized based on the correlation coefficients between the final-level indicators to be optimized and the constraint control variable values; otherwise, the final-level indicators to be optimized in the cluster are optimized based on the constraint control variable values ​​of the final-level indicators to be optimized.

[0053] Preferably, the final index optimization module is used to optimize the final index to be optimized in the cluster based on the correlation coefficient between the final indexes to be optimized and the constraint control variable value, including:

[0054] judge Is it greater than or equal to r0? If so, delete x i 、x j The final indicator to be optimized has a small value of the medium constraint control variable;

[0055] Otherwise, if exist Arrange the first n1 digits in descending order, then keep x i , otherwise delete x i ;

[0056] in,

[0057] x i 、x j are the i-th and j-th final-level indicators to be optimized, respectively, i, j = 1, ..., n; n is the total number of final-level indicators to be optimized;

[0058] For x i With x j The correlation coefficient; r0 is the strong correlation determination threshold;

[0059] For x i With x j The ratio of the values ​​of the constraint control variables; For x i The sum of the ratios of the constraint control variable values ​​of each final level indicator to be optimized in the cluster; m s is the number of final-level indicators to be optimized in the s-th cluster; is the final indicator to be optimized in the same cluster The matrix composed of The calculation formula is:

[0060]

[0061] in, x i 、x j The constraint control variable value is calculated as follows:

[0062]

[0063]

[0064] xi 、x j The relevance of user needs;

[0065] x i 、x j The relevance of technology requirements;

[0066] x i 、x j correction factor.

[0067] Preferably, the final index optimization module is used to optimize the final index to be optimized of the cluster based on the constraint control variable values ​​between the final indexes to be optimized, including:

[0068] like exist If the descending order is the first n2, then keep x i , otherwise delete x i ;

[0069] Among them, x i 、x j are the i-th and j-th final-level indicators to be optimized, respectively, i, j = 1, ..., n; n is the total number of final-level indicators to be optimized;

[0070] For x i With x j The ratio of the values ​​of the constraint control variables; For x i The sum of the ratios of the constraint control variable values ​​of each final-level indicator to be optimized; The final indicator to be optimized The matrix composed of The calculation formula is:

[0071]

[0072] in, x i 、x j The constraint control variable value is calculated as follows:

[0073]

[0074]

[0075] x i 、x j The relevance of user needs;

[0076] xi 、x j The relevance of technology requirements;

[0077] x i 、x j correction factor.

[0078] Preferably, the upper-level indicator optimization module is used to:

[0079] Divide the optimized subordinate indicators into n x clusters, where the ath cluster is represented by X a ; Make new indicator judgments on the subordinate indicators in each cluster and determine the new indicators in each cluster;

[0080] Except for the clusters that all contain the newly added indicators, determine the original superior indicators to which each cluster belongs for the other clusters;

[0081] Determine the ath cluster of the optimized subordinate indicators Is it greater than δ0? If so, define a new parent indicator for the cluster;

[0082] Otherwise, determine the a-th cluster Is it greater than p0? If so, define a new parent indicator for the cluster;

[0083] Otherwise, the original superior indicator to which the cluster belongs is used as the superior indicator of the cluster;

[0084] in, is the optimization control factor of the ath cluster of the optimized subordinate indicators; δ0 is the optimization control judgment threshold; is the proportion of newly added indicators in the ath cluster of the optimized subordinate indicators; p0 is the judgment threshold of the newly added indicators.

[0085] Beneficial effects of the present invention:

[0086] (1) The present invention includes the optimization of the final level index and the upper level index, which is suitable for the optimization of a multi-level index system and is more comprehensive and reasonable.

[0087] (2) The final index optimization method of the present invention can more comprehensively cover three different types of index deletion, addition and replacement.

[0088] (3) The final indicator optimization model in the present invention takes into account the characteristic relationship between the new indicators and the original indicators, the clustering relationship and the strength of the correlation of the final indicators, the correlation between the final indicators and user needs, the correlation with technical needs, and the influence of other factors such as mandatory regulatory indicators, so that the optimization path is more scientific and reasonable.

[0089] (4) The superior indicator optimization model in the present invention takes into account the affiliation of the original superior and subordinate indicators, the clustering relationship of the optimized subordinate indicators, the relationship between the original superior indicators and the clustering of the optimized subordinate indicators, and other factors such as the influence of the proportion of new indicators, thus filling the gap in the superior indicator optimization path. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] Figure 1 A schematic diagram of the superior indicator optimization strategy provided by the present invention;

[0091] Figure 2 A schematic diagram of the final-level indicator optimization strategy provided by the present invention. DETAILED DESCRIPTION

[0092] The present invention is further described below in conjunction with the embodiments. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of the present invention.

[0093] The present invention provides a demand-driven product evaluation index system optimization method, comprising the following steps:

[0094] (1) Determine the final-level indicators to be optimized and build a final-level indicator library to be optimized. The final-level indicators to be optimized include the original final-level indicators and the newly added final-level indicators. The newly added final-level indicators are determined based on user needs, quality feedback, R&D needs, and standards and regulations.

[0095] (2) Perform cluster analysis, correlation analysis and new indicator judgment on the final indicators to be optimized to obtain the relationship characteristics between the final indicators to be optimized, including:

[0096] Define x i 、x j are the i-th and j-th final-level indicators to be optimized, respectively, i, j = 1, ..., n; n is the total number of final-level indicators to be optimized;

[0097] The final level indicators to be optimized contain a total of m clusters, among which the number of final level indicators to be optimized in the sth cluster is m s ;

[0098] A new indicator is judged for the final indicator to be optimized in each cluster to determine whether each final indicator to be optimized is a new final indicator;

[0099] Perform correlation analysis on the final indicators to be optimized in each cluster and obtain x i 、x j The correlation coefficient

[0100] (3) Calculate the constraint control variable value of the final indicator to be optimized and construct the constraint control matrix of the final indicator to be optimized

[0101]

[0102] in,

[0103] G. is the constraint control matrix of the final indicator to be optimized;

[0104] For x i 、x j The ratio of the values ​​of the constraint control variables; The calculation formula is:

[0105]

[0106] x i 、x j The constraint control variable value is calculated as follows:

[0107]

[0108]

[0109] in,

[0110] For x i 、x j The relevance of user needs, It can be obtained through user surveys. The survey questions are in the form of correlation scores (1-10 points). According to the survey results, the correlation of user needs is set within (0, 1);

[0111] For x i 、x j The correlation between technical requirements, It can be obtained through technical personnel survey. The survey questions are in the form of scoring the degree of technical relevance (1-10 points). According to the survey results, the user demand relevance is set within (0, 1);

[0112] For x i 、x j The correction factor is determined based on other influencing factors that determine whether the final indicator to be optimized is retained, including mandatory regulatory requirements, whether it is a new final indicator, etc. The correction factor can be the product of the correction factors represented by these influencing factors. If it is a mandatory regulatory requirement, the final indicator to be optimized should be retained, and the correction factor value is larger; if it is a new indicator, the correction factor is greater than 1, and if it is not a new indicator, the correction factor is less than 1.

[0113] (4) Optimizing the final indicators

[0114] See also Figure 2 , if there are multiple final indicators to be optimized in the sth cluster, then judge x i 、x j The correlation coefficient Is it greater than or equal to the strong correlation determination threshold r0? If so, then x i 、x j Has a strong correlation, retain x i 、x j The final level indicator to be optimized with a large value of the constraint control variable is deleted, and the final level indicator to be optimized with a small value of the constraint control variable is deleted; if is less than the strong correlation determination threshold r0, then x i 、x j Weak correlation, x i Compared with other final indicators to be optimized in the same cluster exist Arrange the first n1 bits in descending order, and retain the final indicator to be optimized; otherwise, delete the final indicator to be optimized.

[0115] in, For x i The sum of the ratios of the constraint control variable values ​​of each final-level indicator to be optimized in the cluster;

[0116] is the final indicator to be optimized in the sth cluster The matrix composed of

[0117]

[0118] If there is only one final indicator x to be optimized in the sth cluster i , x i and other final indicators to be optimized exist If the index is in the first n2 positions in descending order, the last-level index to be optimized is retained; otherwise, the last-level index to be optimized is deleted;

[0119] in, For x i The sum of the ratios of the constraint control variable values ​​of each final-level indicator to be optimized; The final indicator to be optimized The matrix composed of

[0120] Generally, n2 is greater than the number of clusters whose index number is 1 in the clusters of the final-level index to be optimized.

[0121] In this embodiment, the final target is optimized by using the final target constraint control variable value expected large characteristic. In other embodiments of the present invention, the final target can also be optimized by using the final target constraint control variable value expected small characteristic or expected target characteristic.

[0122] (5) Perform cluster analysis on the optimized lower-level indicators and divide them into n x clusters, where the ath cluster is represented by X a ; Make new indicator judgments on the subordinate indicators in each cluster and determine the new indicators in each cluster;

[0123] (6) Performing affiliation analysis on the optimized lower-level indicators and the original upper-level indicators. Except for the clusters that all contain the newly added indicators, the original upper-level indicators to which each cluster belongs are determined for the other clusters; one cluster belongs to one upper-level indicator;

[0124] (7) Optimize the superior indicators

[0125] If the optimization control factor of the ath cluster of the lower-level index after optimization If it is greater than the optimization control judgment threshold δ0, a new superior indicator is defined for the cluster. is the optimization control factor of the ath cluster after optimization. Its value is not unique and is determined by the factors that affect whether the upper-level indicators in the ath cluster are retained. The influencing factors are whether the newly added final-level indicators contained in the ath cluster contain the newly added mandatory indicators. If the ath cluster contains the newly added mandatory indicators, The meaning is that the cluster needs to define a new upper-level indicator to represent the lower-level indicator cluster. The newly added mandatory requirement means that the newly added lower-level indicator included in the a-th cluster comes from the mandatory regulatory requirements.

[0126] Otherwise, compare the proportion of new indicators in the ath cluster If the threshold p0 of the newly added indicator is greater than p0, a new upper-level indicator is defined for the cluster. The threshold p0 of the newly added indicator is a constant. It is the proportion of newly added indicators in the a-th cluster, which refers to the ratio of the number of newly added indicators in the a-th cluster to the total number of indicators in the cluster.

[0127] Otherwise, the original superior indicator to which the cluster belongs is used as the superior indicator of the cluster.

[0128] That is, in each cluster of optimized subordinate indicators, if the similarity with the original superior indicator is high, the name of the original superior indicator is used and an explanation is given; if the similarity is low or the subordinate indicator contains a newly added mandatory indicator, a new superior indicator name is redefined for the cluster and a new explanation is given. At this time, the original superior indicator to which the cluster belongs is replaced; if the cluster is all new indicators, there is no superior indicator corresponding to the cluster in the original superior indicator. It is necessary to define the superior indicator name for the cluster and give a new explanation. At this time, the superior indicator to which the cluster belongs is a new indicator. After the optimization of subordinate indicators, there are situations such as the addition and deletion of indicators. The original subordinate indicator affiliation relationship is no longer applicable or is not accurate enough. Figure 1 The upper-lower and lower-level optimization strategy re-determines the upper-level indicator to which each cluster of the optimized lower-level indicators belongs, and completes the optimization of the upper-level indicators. After the optimization, the number of upper-level indicators is equal to the number of clusters of the lower-level indicators.

[0129] (8) Repeat steps (5) to (7) to optimize the upper-level indicators step by step until the optimization of indicators at all levels is completed.

[0130] The embodiment of the present invention also provides a demand-driven product evaluation index system optimization model, including a final index optimization module and a superior index optimization module;

[0131] The final index optimization module is used to determine the final index to be optimized; perform cluster analysis and correlation analysis on the final index to be optimized, and calculate the constraint control variable value of the final index to be optimized; optimize the final index based on the cluster relationship, correlation coefficient and constraint control variable value between the final indexes to be optimized;

[0132] The upper-level indicator optimization module is used to perform cluster analysis on the optimized lower-level indicators; perform affiliation analysis on the optimized lower-level indicators and the original upper-level indicators; and optimize the upper-level indicators step by step based on the optimization control factors of the lower-level indicators of the same cluster in the optimized lower-level indicators and the proportion of new indicators until all indicators are optimized.

[0133] The final index optimization module is used to determine whether the number of final indexes to be optimized in the same cluster is greater than 1. If so, the final indexes to be optimized in the cluster are optimized based on the correlation coefficients between the final indexes to be optimized and the values ​​of the constraint control variables, including:

[0134] judge Is it greater than or equal to r0? If so, delete x i 、x j The final indicator to be optimized has a small value of the medium constraint control variable;

[0135] Otherwise, if exist Arrange the first n1 digits in descending order, then keep xi , otherwise delete x i ;

[0136] in,

[0137] x i 、x j are the i-th and j-th final-level indicators to be optimized, respectively, i, j = 1, ..., n; n is the total number of final-level indicators to be optimized;

[0138] For x i With x j The correlation coefficient; r0 is the strong correlation determination threshold;

[0139] For x i With x j The ratio of the values ​​of the constraint control variables; For x i The sum of the ratios of the constraint control variable values ​​of each final level indicator to be optimized in the cluster; m s is the number of final-level indicators to be optimized in the s-th cluster; is the final indicator to be optimized in the same cluster The matrix composed of The calculation formula is:

[0140]

[0141] in, x i 、x j The constraint control variable value is calculated as follows:

[0142]

[0143]

[0144] For x i 、x j The relevance of user needs, It can be obtained through user surveys. The survey questions are in the form of correlation scores (1-10 points). According to the survey results, the correlation of user needs is set within (0, 1);

[0145] For x i 、x j The correlation between technical requirements, It can be obtained through technical personnel survey. The survey questions are in the form of scoring the degree of technical relevance (1-10 points). According to the survey results, the user demand relevance is set within (0, 1);

[0146] For x i 、x j The correction factor is determined based on other influencing factors that determine whether the final indicator to be optimized is retained, including mandatory regulatory requirements, whether it is a new final indicator, etc. The correction factor can be the product of the correction factors represented by these influencing factors. If it is a mandatory regulatory requirement, the final indicator to be optimized should be retained, and the correction factor value is larger; if it is a new indicator, the correction factor is greater than 1, and if it is not a new indicator, the correction factor is less than 1.

[0147] Otherwise, the final level indicator to be optimized of the cluster is optimized based on the constraint control variable value of the final level indicator to be optimized, including:

[0148] like exist If the descending order is the first n2, then keep x i , otherwise delete x i ;

[0149] For x i The sum of the ratios of the constraint control variable values ​​of each final-level indicator to be optimized; The final indicator to be optimized The matrix composed of.

[0150] The upper-level indicator optimization module is used to:

[0151] Divide the optimized subordinate indicators into n x clusters, where the ath cluster is represented by X a ; Make new indicator judgments on the subordinate indicators in each cluster and determine the new indicators in each cluster;

[0152] Except for the clusters that all contain the newly added indicators, determine the original superior indicators to which each cluster belongs for the other clusters;

[0153] Determine the ath cluster of the optimized subordinate indicators Is it greater than δ0? If so, define a new parent indicator for the cluster;

[0154] Otherwise, determine the a-th cluster Is it greater than p0? If so, define a new parent indicator for the cluster;

[0155] Otherwise, the original superior indicator to which the cluster belongs is used as the superior indicator of the cluster;

[0156] in, is the optimization control factor of the ath cluster of the optimized subordinate indicators; δ0 is the optimization control judgment threshold; is the optimization control factor of the ath cluster after optimization. Its value is not unique and is determined by the factors that affect whether the upper-level indicators in the ath cluster are retained. The influencing factors are whether the newly added final-level indicators contained in the ath cluster contain the newly added mandatory indicators. If the ath cluster contains the newly added mandatory indicators, The meaning is that the cluster needs to define a new upper-level indicator to represent the lower-level indicator cluster. The newly added mandatory requirement means that the newly added lower-level indicator included in the a-th cluster comes from the mandatory regulatory requirements.

[0157] is the proportion of newly added indicators in the ath cluster of the optimized subordinate indicators; p0 is the judgment threshold of the newly added indicators.

[0158] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0159] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0160] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0161] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0162] 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 purpose of the present invention and the claims, which all fall within the protection of the present invention.

[0163] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A demand-driven product evaluation index system optimization method, characterized in that: include, S1: Determine the final level indicator to be optimized; perform cluster analysis and correlation analysis on the final level indicator to be optimized, and calculate the constraint control variable value of the final level indicator to be optimized; optimize the final level indicator based on the cluster relationship, correlation coefficient and constraint control variable value between the final level indicators to be optimized; S2: cluster analysis of optimized subordinate indicators; Conduct affiliation analysis on the optimized subordinate indicators and the original superior indicators; optimize the superior indicators based on the optimized control factors of the subordinate indicators of the same cluster and the proportion of the newly added indicators in the optimized subordinate indicators; S3: Repeat step S2 to optimize the upper-level indicators step by step until all indicators are optimized; The final level indicators to be optimized include the original final level indicators and the newly added final level indicators; The cluster analysis of the optimized lower-level indicators includes: dividing the optimized lower-level indicators into n x clusters, where the ath cluster is represented by X a ; Make new indicator judgments on the subordinate indicators in each cluster and determine the new indicators in each cluster; The subordination analysis of the optimized lower-level indicators and the original upper-level indicators includes: determining the original upper-level indicators to which each cluster belongs for the other clusters except for the clusters that all contain the newly added indicators; The optimization of the upper-level indicators based on the optimization control factors of the lower-level indicators of the same cluster and the proportion of the newly added indicators in the optimized lower-level indicators includes: Determine the ath cluster of the optimized subordinate indicators Is it greater than δ0? If so, define a new parent indicator for the cluster; Otherwise, determine the a-th cluster Is it greater than p0? If so, define a new parent indicator for the cluster; Otherwise, the original superior indicator to which the cluster belongs is used as the superior indicator of the cluster; in, is the optimization control factor of the ath cluster of the optimized subordinate indicators; δ0 is the optimization control judgment threshold; is the proportion of newly added indicators in the ath cluster of the optimized subordinate indicators; p0 is the judgment threshold of the newly added indicators; The optimization of the final level indicators based on the clustering relationship, correlation coefficient and constraint control variable value between the final level indicators to be optimized includes: Determine whether the number of final-level indicators to be optimized in the same cluster is greater than 1. If so, optimize the final-level indicators to be optimized in the cluster based on the correlation coefficients between the final-level indicators to be optimized and the values ​​of the constraint control variables; otherwise, optimize the final-level indicators to be optimized in the cluster based on the values ​​of the constraint control variables of the final-level indicators to be optimized; The optimizing the final level index to be optimized in the cluster based on the correlation coefficient between the final level indexes to be optimized and the constraint control variable value includes: judge Is it greater than or equal to r0? If so, delete x i 、x j The final indicator to be optimized has a small value of the medium constraint control variable; Otherwise, if exist Arrange the first n1 digits in descending order, then keep x i , otherwise delete x i ; in, x i 、x j are the i-th and j-th final-level indicators to be optimized, respectively, i, j = 1, ..., n; n is the total number of final-level indicators to be optimized; For x i With x j The correlation coefficient; r0 is the strong correlation determination threshold; For x i With x j The ratio of the values ​​of the constraint control variables; For x i The sum of the ratios of the constraint control variable values ​​of each final level indicator to be optimized in the cluster; m s is the number of final-level indicators to be optimized in the s-th cluster; is the final indicator to be optimized in the same cluster The matrix composed of The calculation formula is: in, x i 、x j The constraint control variable value is calculated as follows: x i 、x j The relevance of user needs; x i 、x j The relevance of technology requirements; x i 、x j Correction factor of The method of optimizing the final level index to be optimized of the cluster based on the constraint control variable value between the final level indexes to be optimized includes: like exist If the descending order is the first n2, then keep x i , otherwise delete x i ; in, For x i The sum of the ratios of the constraint control variable values ​​of each final-level indicator to be optimized; The final indicator to be optimized The matrix composed of.

2. The method for optimizing a demand-driven product evaluation index system according to claim 1, characterized in that: Determining the final level indicator to be optimized includes: Determine new final-level indicators based on user needs, quality feedback, R&D needs, and standards and regulations.

3. A demand-driven product evaluation index system optimization model, characterized in that: It includes the final index optimization module and the upper index optimization module; The final-level indicator optimization module is used to determine the final-level indicators to be optimized, which include the original final-level indicators and the newly added final-level indicators; perform cluster analysis and correlation analysis on the final-level indicators to be optimized, and calculate the constraint control variable values ​​of the final-level indicators to be optimized; optimize the final-level indicators based on the clustering relationship, correlation coefficient and constraint control variable values ​​between the final-level indicators to be optimized; The upper-level indicator optimization module is used to perform cluster analysis on the optimized lower-level indicators; Perform affiliation analysis on the optimized subordinate indicators and the original superior indicators; optimize the superior indicators based on the optimization control factors of the subordinate indicators of the same cluster and the proportion of the newly added indicators in the optimized subordinate indicators; optimize the superior indicators step by step until all indicators are optimized; The cluster analysis of the optimized lower-level indicators includes: dividing the optimized lower-level indicators into n x clusters, where the ath cluster is represented by X a ; Make new indicator judgments on the subordinate indicators in each cluster and determine the new indicators in each cluster; The subordination analysis of the optimized lower-level indicators and the original upper-level indicators includes: determining the original upper-level indicators to which each cluster belongs for the other clusters except for the clusters that all contain the newly added indicators; The optimization of the upper-level indicators based on the optimization control factors of the lower-level indicators of the same cluster and the proportion of the newly added indicators in the optimized lower-level indicators includes: Determine the ath cluster of the optimized subordinate indicators Is it greater than δ0? If so, define a new parent indicator for the cluster; Otherwise, determine the a-th cluster Is it greater than p0? If so, define a new parent indicator for the cluster; Otherwise, the original superior indicator to which the cluster belongs is used as the superior indicator of the cluster; in, is the optimization control factor of the ath cluster of the optimized subordinate indicators; δ0 is the optimization control judgment threshold; is the proportion of newly added indicators in the ath cluster of the optimized subordinate indicators; p0 is the judgment threshold of the newly added indicators; The optimization of the final level indicators based on the clustering relationship, correlation coefficient and constraint control variable value between the final level indicators to be optimized includes: Determine whether the number of final-level indicators to be optimized in the same cluster is greater than 1. If so, optimize the final-level indicators to be optimized in the cluster based on the correlation coefficients between the final-level indicators to be optimized and the values ​​of the constraint control variables; otherwise, optimize the final-level indicators to be optimized in the cluster based on the values ​​of the constraint control variables of the final-level indicators to be optimized; The optimizing the final level index to be optimized in the cluster based on the correlation coefficient between the final level indexes to be optimized and the constraint control variable value includes: judge Is it greater than or equal to r0? If so, delete x i 、x j The final indicator to be optimized has a small value of the medium constraint control variable; Otherwise, if exist Arrange the first n1 digits in descending order, then keep x i , otherwise delete x i ; in, x i 、x j are the i-th and j-th final-level indicators to be optimized, respectively, i, j = 1, ..., n; n is the total number of final-level indicators to be optimized; For x i With x j The correlation coefficient; r0 is the strong correlation determination threshold; For x i With x j The ratio of the values ​​of the constraint control variables; For x i The sum of the ratios of the constraint control variable values ​​of each final level indicator to be optimized in the cluster; m s is the number of final-level indicators to be optimized in the s-th cluster; is the final indicator to be optimized in the same cluster The matrix composed of The calculation formula is: in, x i 、x j The constraint control variable value is calculated as follows: x i 、x j The relevance of user needs; x i 、x j The relevance of technology requirements; x i 、x j Correction factor of The method of optimizing the final level index to be optimized of the cluster based on the constraint control variable value between the final level indexes to be optimized includes: like exist If the descending order is the first n2, then keep x i , otherwise delete x i ; in, For x i The sum of the ratios of the constraint control variable values ​​of each final-level indicator to be optimized; The final indicator to be optimized The matrix composed of.

Images