JPH0749883A - Computer - Google Patents

Abstract

PURPOSE:To speed up and facilitate quantitative evaluation by detecting a factor which affects the degree of user's satisfaction in performance from the performance results of respective measured functions and evaluating the performance by using an evaluation expression generated in consideration of the degree of user's satisfaction. CONSTITUTION:When performance data are inputted, factor scores of the respective functions are calculated by a factor analyzing process 210 by using stored evaluation setting data. Then a multiple recurrent analyzing process is performed in consideration of both the factor scores and the degree of user's satisfaction in the performance and the obtained recurrent expression is stored as the performance evaluation expression for which the degree of user's satisfaction is taken into consideration. Then an evaluation calculating means 220 substitutes the factor scores in the evaluation expression by using the stored evaluation expression data to calculate evaluated values for the respective functions. The evaluation results are displayed on an evaluation result display screen and also registered in a performance evaluation list. Consequently, the performance can automatically be evaluated in consideration of the degree of user' satisfaction.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a computer system capable of easily evaluating the performance of each function of information equipment such as word processors and personal computers.

[0002]

2. Description of the Related Art With the spread of information equipment, in the development of those products, appropriate performance evaluation is performed on conventional equipment and prototypes of the equipment to be developed, and functions and processing that require performance improvement are extracted. Taking prompt performance measures against them is an important point in developing good products.

Conventionally, as a performance evaluation method, a benchmark test or the like is used to measure the performance of each function and compare it with the performance of a conventional machine or another company's machine, or a user's evaluation of the performance by a questionnaire or a hearing survey. The method of obtaining directly is mainly used. In addition, using a performance measurement tool (Japanese Patent Laid-Open No. 02-139644), measure the processing time and frequency of use of tasks and supervisor calls in each function, and extract the processing part that is the bottleneck. Is mentioned.

[0004]

When evaluating the performance of a device, it is necessary to consider whether the performance satisfies the user (satisfaction level of the user). The conventional method has the following problems.

(1) The evaluation method using a benchmark test or the like can easily and quickly perform the performance evaluation, but since only the performance between the devices is compared, the satisfaction of the user is taken into consideration. Not not.

(2) Evaluation methods such as questionnaires and hearing surveys can directly obtain user satisfaction with performance, but it takes time to collect and organize data.

(3) In the evaluation method using the performance measurement tool, it is possible to easily grasp the time-consuming part in the processing of each function, but the evaluation is performed by focusing only on the measured performance value. Therefore, the satisfaction of the user is not taken into consideration as in (1).

In order to solve the above problems, the object of the present invention is as follows.

There is provided a computer system having a function capable of quickly and easily performing quantitative evaluation of performance in consideration of user's satisfaction by using the performance of equipment.

[0010]

[Means for Solving the Problems] The above-mentioned purpose is a factor analysis method which is one method of statistical analysis (reference: Kyoritsu Publishing, PC Statistical Analysis Handbook-II Multivariate Analysis, pp.160-16).
3, 195-200), from the performance results of each function measured using benchmark tests and performance measurement tools,
Extract factors that affect user satisfaction with performance,
Using the multiple regression analysis method (reference: Kyoritsu Publishing, PC Statistical Analysis Handbook-II Multivariate Analysis, pp. 1-7),
The result of the user's evaluation of each function by the factor and the questionnaire survey (that is, the user's satisfaction)
This is achieved by additionally performing multiple regression analysis and using the regression equation derived from the result of the regression analysis as an evaluation equation in consideration of user satisfaction.

The means for performance evaluation using the evaluation formula will be described below.

(1) Performance Data Input Means The performance data input means allows the user to input and set the performance for each required function of the device when performing the performance evaluation using the above-described evaluation formula. The means necessary for that purpose are as follows.

Performance evaluation instruction means for the user to perform performance evaluation Display a screen (performance data input display screen) for the user to input and set the response time (hereinafter abbreviated as performance data) of each function required at the time of evaluation Performance data input display screen display means Performance data input means on which the user inputs performance data on the performance data input display screen (2) Performance evaluation means (1) Factors using performance data input by the performance data input means An analysis is performed to quantitatively express the factors that affect the user's satisfaction with the performance, and the factors are substituted into the evaluation formula in consideration of the user's satisfaction. Then, the evaluation value for each function is calculated. The necessary means for this are as follows.

Evaluation setting data storage means for storing data (for example, types of explanatory variables, the number of factors, etc., abbreviated as evaluation setting data hereinafter) necessary for performing factor analysis during performance evaluation (1) Performance input means Using the input performance data and evaluation setting data, in order to express the factor for each function quantitatively (hereinafter abbreviated as factor score), factor analysis means for factor analysis (3) Evaluation value calculation means for calculating an evaluation value for each function by substituting a factor score obtained from the factor analysis result of the evaluation expression data storage means for storing expression data) into the evaluation expression using the evaluation expression data (3) ) Evaluation result output means The means for displaying the evaluation result to the user is as follows.

(2) Evaluation result display means for displaying the evaluation value for each function obtained by the performance evaluation means on the screen (evaluation result display screen) (4) Evaluation result registration means Means for registering the evaluation result are as follows. Is the street.

Performance evaluation list storing performance data and evaluation results for each function Evaluation result registration instruction means for the user to register the evaluation results in the performance evaluation list Evaluation in which the user inputs a file name for registering the evaluation results Evaluation result registration screen display means for displaying a result registration screen Evaluation result registration means for registering performance data and performance evaluation results in the performance evaluation list By the means (1) to (4), the satisfaction level of the user using the evaluation formula It is possible to perform performance evaluation in consideration of.

Regarding the evaluation formula, there are a method of setting the evaluation formula in advance in the system as a default and a method of allowing the user to newly derive (change) the evaluation formula by providing the following means (5). Conceivable.

(5) Evaluation formula deriving means The means for the user to newly derive (change) the evaluation formula is as follows.

Evaluation formula derivation instructing means for the user to derive the evaluation formula As a result of investigating the actual satisfaction of the user with respect to the performance data and the performance required when performing the above-described factor analysis and multiple regression analysis (hereinafter, (Regression analysis data) Input the regression analysis data using the regression analysis data input means to perform the multiple regression analysis in order to derive the evaluation equation. Evaluation formula registration means for registering coefficient and constant terms, etc.) in the evaluation formula data

[0020]

When the user instructs the performance evaluation of the device by the performance evaluation instruction means, the performance data input display screen display means displays the performance data input display screen. Then, the performance data input means prompts the user to input performance data using a keyboard or the like.

When the performance data is input, the factor analysis means calculates the factor score for each function using the evaluation setting data stored in the evaluation setting data storage means. Then, the evaluation value calculation unit calculates the evaluation value for each function by substituting the factor score into the evaluation formula using the evaluation formula data stored in the evaluation formula data storage unit.

The evaluation result is displayed on the evaluation result display screen by the evaluation result display means.

When the evaluation result registration instruction means instructs the user to register the evaluation result, the evaluation result registration screen display means displays the evaluation result registration screen, and the user inputs the file name. The data is registered in the performance evaluation list by.

When the evaluation formula derivation instructing unit instructs the user to derive (change) the evaluation formula, the regression analysis data input unit prompts the user to input regression analysis data for deriving the evaluation formula. Then, using the regression analysis data, an evaluation formula is derived by multiple regression analysis means. Further, the evaluation formula registration means registers data relating to the evaluation formula in the evaluation formula data.

As described above, in the computer system, the user can automatically perform the performance evaluation in consideration of the satisfaction of the user by inputting only the performance data at the time of the performance evaluation.

[0026]

【Example】

(Embodiment 1 of Means (1) to (4)) Hereinafter, embodiments of the present invention will be described with reference to the drawings.

First, the hardware configuration of the information processing apparatus according to the embodiment of the present invention will be described with reference to FIG.
The information processing device 1 includes a central processing unit (CPU) 10, a main memory (MM) 11, and a video memory (VRAM) 12.
, Keyboard (KB) 13 and mouse (MS) 14
, A display device (CRT) 15, a fixed disk device (HDD) 16, and a floppy disk (FDD) 1
7 is provided.

The central processing unit (CPU) 10 executes various programs for performance evaluation stored in the main memory (MM) 11 and controls the peripheral devices 11 to 17. C
The content of the display data stored in the video memory (VRAM) 12 by the PU 10 is displayed on the display device (CRT) 15. The keyboard (KB) 1 is used for performance evaluation, input of data required for evaluation, instruction input such as registration of evaluation results, etc.
3 and mouse (MS) 14. A fixed disk device (HDD) 16 reads or saves various programs and data relating to performance evaluation. Data transfer between these peripheral devices 11 to 17 and the CPU 10 is performed via the bus 18.

Next, the outline of the performance evaluation procedure in this system will be described with reference to FIG.

First, in the present system, when the user inputs the performance for each function of the device (hereinafter, abbreviated as performance data) necessary for performance evaluation (step 100), the performance evaluation is performed using the performance data. Perform processing (step 20)
0). Then, the result of the performance evaluation is displayed on the screen (step 300), and the evaluation result is registered in a file (hereinafter, performance evaluation list) (step 400). Details of the procedure are described below.

First, the structure of evaluation setting data for storing variables required for performance evaluation will be described with reference to FIG.

The evaluation setting data 500 includes the number of types of performance data 510, the name of evaluation target 520, the name of comparison target 530, the number of factors 540 when performing factor analysis described later, and the number of function types 55.
0, 56 types of explanatory variables when performing factor analysis described later
0, an evaluation formula data pointer 570.

The number of types of performance data 510 is the number of types of equipment (two or more) required for performance evaluation, and the performance data of the equipment is input by the user at the time of performance evaluation. The number of types of performance data 510 is a value preset in the present system, and is stored in the fixed disk device 16.

The evaluation target name 520 is a name (one type) of a device to be evaluated (hereinafter, abbreviated as an evaluation target device) among the performance data type numbers 510.

The comparison target 530 is the number 5 of types of performance data.
Name (1 or more) of the device to be compared with the performance of the device to be evaluated (hereinafter abbreviated as the device to be compared) out of 10.
Therefore, these performances are considered in the performance evaluation. This is because the satisfaction level of the user with respect to the performance largely depends on the performance of the device used by the user in the past, so that it is necessary to consider not only the device to be evaluated but also the performance of other devices. .

The number of factors 540 is the number of factors required for performing factor analysis at the time of performance evaluation (step 200), which will be described later, and is a value preset in this system and stored in the fixed disk device 16. Has been done.

The number of function types 550 is the number of types of functions (for example, character deletion, copying, movement, etc.) of the device to be evaluated, and is a value preset in this system, and is set in the fixed disk device 16. It is stored.

The number of kinds of explanatory variables 560 is the number of kinds of variables (hereinafter, abbreviated as explanatory variables) necessary for performing factor analysis at the time of performance evaluation (step 200) described later, and is set in advance in this system. Value stored in the fixed disk device 16.

The evaluation formula data pointer 570 stores a pointer indicating a position where data relating to an evaluation formula used at the time of performance evaluation described later is stored.

The configuration of the performance data input display screen in which the user inputs performance data will be described with reference to FIG.

On the performance data input / display screen 600, the first screen 610 for the user to input the type name of the device required for performance evaluation, and the first screen 610 for the input of the performance data for each function of the device. A second screen 650 is displayed.

On the first screen 610, the name 620 of the equipment to be evaluated and the name 630 of the equipment to be compared, which are necessary for the performance evaluation, and the names 620 and 630, are input and first execution is selected. Key 640 is displayed. On the second screen 650, the function 66 of the device to be evaluated at the time of performance evaluation.
0, the performance data 670 for each function 660 of the device indicated by the name 620, the performance data 680 for each function 660 of the device indicated by the name 630, and the second selected after inputting the performance data 670, 680. The execute key 690 is displayed.

Next, the configuration of the evaluation result display screen for displaying the performance evaluation result will be described with reference to FIG. On the evaluation result display screen 700, the evaluation values 71 obtained by sorting the evaluation values obtained as a result of performance evaluation in descending or ascending order.
0, a function 720 corresponding thereto, and a registration instruction key 730 for registering the evaluation result are displayed.

Next, the configuration of the evaluation result registration screen displayed when the performance evaluation result is registered (step 400) will be described with reference to FIG. On the evaluation result registration screen 800, a file name 810 and a third execution key 820 to be selected after inputting the file name 810 are displayed, and the user is asked to input the file name to register the evaluation result.

Next, the structure of a file for registering evaluation results (hereinafter, abbreviated as a performance evaluation list) will be described with reference to FIG. The performance evaluation list 900 includes file management data 910 and evaluation result data 920.

The file management data 910 associates the file name with the evaluation result data 920 relating to the file and the evaluation setting data 500. The file management data 910 includes a file name 930 that stores a file name input by the user and the evaluation result data 9
An evaluation result data pointer 940 indicating the location of 20 and an evaluation setting data pointer 950 indicating the location of the evaluation setting data 500.

The evaluation result data 920 includes the function 960 for which the performance is evaluated, the performance data 970 of the device to be evaluated for the function 960, and the performance data 98 of the device to be compared.
0 and an evaluation value 990 for the performance of the evaluation target device.

Next, with reference to FIG. 8, the structure of the evaluation formula data for storing the data relating to the evaluation formula considering the user's satisfaction with the performance will be described. Evaluation formula data 10
00 is a coefficient value (m1, m2, m3 in the equation 7) 1100 of the evaluation formula shown in the equation 7 used in the performance evaluation described later.
Composed of. The coefficient value 1100 is a value preset in this system and is stored in the fixed disk device 16.

Next, the performance data input process 1 shown in FIG.
00 will be described in detail with reference to FIGS. 4 and 9.

First, when the user instructs the performance evaluation of the device using the mouse 14 or the like (step 110), the number of types of performance data stored in the evaluation setting data 500 is 5
10 (2 in FIG. 4) is read (step 120), and the first screen 610 on the performance data input display screen 600 is displayed (step 130). After that, when the user inputs two types of device names 620 and 630 using the keyboard 13 and selects the first execution key 640 using the mouse 14 (step 140), the device name is evaluated and set in the evaluation setting data 500. Evaluation target name 520 and comparison target name 53
Set to 0. Then, the second screen 670 is displayed (step 150), and the performance data 670 of the device name 620 is displayed for each function 660 having the number of function types 550 (50 in FIG. 4) by the user using the keyboard 13.
Then, the performance data 680 of the device name 630 is input, and then the second execution key 690 is selected using the mouse 14 (step 160).

Next, the performance evaluation processing 200 shown in FIG. 2 will be described in detail with reference to FIGS.

First, referring to FIG. 10, the performance evaluation processing 20 is performed.
An outline of 0 is shown.

Factor analysis is performed using the performance data 670 and 680 input by the performance data input processing 100 described above (step 210). Then, the evaluation value Y for each function is calculated using the result of the factor analysis (step 220). Hereinafter, the processing procedure of each step will be described in detail. Details of terms and methods relating to factor analysis are described in Kyoritsu Shuppan, PC Statistical Analysis Handbook-II Multivariate Analysis, pp. 160-163, 195-200.

First, the factor analysis processing 210 will be described with reference to FIGS. 4 and 11.

First, the number of explanatory variable types 560 (3 in FIG. 4) stored in the evaluation setting data 500 is read (step 211). Then, the performance T of the device to be evaluated
1i (the performance data 670 of BC8 in FIG. 4) and the performance T2i of the device to be compared (performance data 680 of BC7 in FIG. 4) are substituted into Equation 1, and three types of explanatory variables f1, f2,
The values f1i, f2i, and f3i for the function i of f3 are calculated (step 212).

Here, in the factor analysis, as explanatory variables,
Set a measurable value that is considered to explain the factors that affect the user's satisfaction with performance. In this embodiment, the value stored in the number of types of performance data 510 is set to “2”, and two types of performance data T1i and T2i are used.
As the main explanatory variables, three types of explanatory variables f1, f2, and f3 shown in Formula 1 were set.

[0057]

[Equation 1]

The explanatory variable f1 is the performance T1i of the device to be evaluated, and it is considered that the magnitude of its absolute value is related to the factor affecting the user's satisfaction with the performance.

The explanatory variable f2 is the performance T1 of the device under evaluation.
i is the difference between the performance T2i of the device to be compared, and if the performance T1i of the device to be evaluated is worse than the performance T2i of the device to be compared, the user's satisfaction will be reduced, thus affecting the user's satisfaction. It is thought to be related to the factors that

The explanatory variable f3 is the ratio of the performance T1i of the device to be evaluated to the performance T2i of the device to be compared. Even if the performance difference between the evaluation target device and the comparison target device (that is, the explanatory variable f2) is the same, if the performance T2i of the comparison target device is different, the performance T of the evaluation target device
Since the ratio (explanatory variable f3) between 1i and the performance T2i of the device to be compared is different, and if this ratio increases, the satisfaction level of the user decreases, so it is considered to be related to the factor affecting the satisfaction level of the user.

From the above, in the present embodiment, the explanatory variable in the case where the device to be compared is one type is set. This is suitable when the performance evaluation is performed in consideration of the satisfaction of only the user who is the device used in the past as the device to be compared. However, in the case of performing performance evaluation in consideration of the satisfaction of a wide range of users, two or more types (other models, different types of information devices, etc.) are set as the devices to be compared, and an explanation including their performance data is given. It is appropriate to newly set a variable in addition to the three types of explanatory variables described above.

[0062]

[Equation 2]

Next, the correlation coefficient rxy between the explanatory variables is calculated using the equation 2 to obtain the correlation matrix C (step 21).
3). Specifically, the correlation coefficient r12 between the explanatory variables f1 and f2
When (r21) is calculated, the average value of f1i and f1i is substituted for xi and xa, the average value of f2i and f2i is substituted for yi and ya, and the number of function types 550 stored in the evaluation setting data 500 is set to N.
(50 in FIG. 4) is substituted. Similarly, f2i and f
Correlation coefficient r23 (r32), r31 (r13) of 3i, f3i and f1i
Is calculated and these values are substituted into the correlation matrix C.

[0064]

[Equation 3]

Next, using the correlation matrix C between the explanatory variables described above, the eigenvalue λ and the eigenvector K corresponding to the eigenvalue λ satisfying the relationship shown in Formula 3 are calculated (step 21).
4). Then, the number of factors 540 (2 in FIG. 4) stored in the evaluation setting data 500 is read (step 21).
5), of the eigenvalues λ, the two with the largest value
Select 1, λ2 (λ1> λ2) and set them as factor 1 and factor 2
And Here, the eigenvectors corresponding to the eigenvalues λ1 and λ2 are K1 and K2.

Next, the factor loading of each factor with respect to the explanatory variable (see PC Statistical Analysis Handbook-II Multivariate Analysis, pp.197) is calculated (step 216). A1 is the factor loading for the explanatory variables f1, f2, and f3 of factor 1.
= (A11, a21, a31), the explanatory variable f1 of factor 2 is
The factor loadings for f2 and f3 are A2 = (a12, a22, a3
2) If the factor loadings A1 and A2 are the eigenvalues λ1 and λ2,
And eigenvectors K1 and K2 can be calculated using Equation 4. Then, from the calculated factor loadings A1 and A2,
Rotation of the factor axis (see PC Statistical Analysis Handbook-II Multivariate Analysis, pp.199), and the factor load after rotation B1 = (b11, b21, b31), B2 = (b12, b2
2, b32) is calculated. Here, in equation 5, P is 3
(The number of explanatory variable types 560) is substituted.

[0067]

[Equation 4]

[0068]

[Equation 5]

Next, a factor score for each function of each factor (PC Statistical Analysis Handbook-II Multivariate Analysis, p.
(see p.200) is calculated (step 217). Factor score Z (factor score Zi1, Z for function i of factor 1 and factor 2)
i2) can be calculated using Equation 6.

[0070]

[Equation 6]

From the above, using the device performances T1i and T2i, the explanatory variable f that is considered to affect the satisfaction of the user
1, f2, f3 were calculated, and the correlation matrix C of the explanatory variables was calculated. Then, regarding the common factors behind the explanatory variables extracted from the correlation matrix C, that is, factor 1 and factor 2,
Factor scores Zi1 and Zi2 for each function were calculated.

Then, the calculated factor score Z is substituted into the evaluation formula (Equation 7) to calculate the evaluation value Yi for each function i (evaluation value calculation 220). The evaluation value Yi is an evaluation value finally obtained as a result of performance evaluation in consideration of user satisfaction. Here, the coefficient m of the evaluation formula (Equation 7)
The values stored in the evaluation formula data 1000 are substituted for 1, m2, and m3. Details of the coefficients m1, m2 and m3 will be described in the second embodiment.

[0073]

[Equation 7]

Next, the evaluation result output process 30 shown in FIG.
0 will be described with reference to FIG.

First, the evaluation values Yi for the respective functions i calculated as described above are sorted in descending order (or ascending order) (step 310), and the sorted evaluation values Yi are displayed on the evaluation result display screen 700 (step 320). .

Next, the evaluation result registration processing 40 shown in FIG.
0 will be described with reference to FIG.

First, when the user instructs the registration of the evaluation result by selecting the registration instruction key 730 with the mouse 14 on the evaluation result display screen 700 (step 410), the evaluation result registration screen 800 is displayed. (Step 420). When the user inputs a file name on the evaluation result registration screen 800 using the keyboard 13 (step 430), the performance data T1i, T2i and the evaluation value Yi are registered (step 440).

Next, the evaluation value registration processing 440 will be described with reference to FIG. The file name input in the file name input process 430 is stored in the file name 930, and the performance data T1i, T2i and the evaluation value Yi are stored in the evaluation result data 920. Then, the evaluation result data 920
Is the evaluation result data pointer 9
Evaluation setting data 5 stored in 40 and used during performance evaluation
The head pointer indicating the location of 00 is stored in the evaluation setting data pointer 950.

As described above, in the present embodiment, when the performance evaluation of the device is performed in consideration of the user's satisfaction, the data required by the user during the performance evaluation, that is, the names of the evaluation target device and the comparison target device ( 620, 630) and the performance data (670, 680) for each function are input to the system, and the factor analysis processing 210 is performed using the above data.
Then, a factor score Z for a factor that influences the user's satisfaction with respect to performance is calculated for each function, and the evaluation value Yi is calculated by substituting the factor score Z into an evaluation formula considering the user's satisfaction, The evaluation value Yi is output as a value that quantitatively represents the user's satisfaction with the performance.

Therefore, by using the evaluation method described in this embodiment, it is possible to easily and quickly evaluate the performance of the device in consideration of the satisfaction of the user.

(Embodiment 2 of Means (5)) Embodiment 1
Then, the evaluation setting data 500 required for performance evaluation
(Evaluation formula data 1000 (coefficient value 1100) that stores data related to (evaluation data type number 510, factor number 540, function type number 550, explanatory variable type number 560) and evaluation formula type
Was described as a default setting in this system in advance. In this case, an evaluation formula in consideration of the satisfaction of the user used by the user in performance evaluation is derived in advance,
The data required for performance evaluation (evaluation setting data 50
0 and the evaluation formula data 1000) are stored in the fixed disk device 16 of the present system. Therefore, when changing the evaluation formula, the user newly derives the evaluation formula and stores the performance evaluation stored in the fixed disk device 16. Sometimes it is necessary to (manually) change the data needed.

Therefore, in the present embodiment, the evaluation formula is derived (changed) only by having the user input the data necessary for deriving (changing) the evaluation formula.
However, a means for automatically setting (changing) the data stored in the fixed disk device 16 and required at the time of performance evaluation will be described below.

First, the structure of the evaluation formula data used in this embodiment will be described with reference to FIG. The evaluation formula data 2000 is the coefficients m1, m2, m in the evaluation formula (Equation 7).
Evaluation formula setting data 2100 for storing values such as 3 and regression analysis data 2200 for storing data necessary for deriving the coefficients m1, m2, m3 and the like.

The evaluation formula setting data 2100 is composed of a regression analysis data pointer 2110 and a coefficient value 2120. The regression analysis data pointer 2110 stores a pointer indicating the location of the regression analysis data 2200. The coefficient value 2120 is the coefficient value m1, m2, m of the derived evaluation formula.
Store 3rd magnitude.

The regression analysis data 2200 includes the function type 22
10, device performance data Til2220, performance data Ti22
230 and dependent variable 2240. Function type 2
Reference numeral 210 denotes a device function required when deriving (changing) the evaluation formula and performing performance evaluation. Performance data Til2220
Indicates the performance of each function 2210 of the device (one type) to be evaluated, which is required when deriving (changing) the evaluation formula. The performance data Ti22230 indicates the performance of each function 2210 of the device (one or more types) to be compared, which is required when deriving (changing) the evaluation formula. The dependent variable 2240 is data used in a multiple regression analysis process described later, and indicates the user's actual evaluation of the performance (performance data Til2220) of the evaluation target device (for example, a survey result of a questionnaire or an interview).

Next, the evaluation formula derivation process 3 will be described with reference to FIG.
000 will be described.

When the user instructs the derivation of the evaluation formula using the mouse 14 or the like (step 3100), the user uses the keyboard 13 to enter the regression analysis data 2200.
(Function type 2210, performance data Til of the device to be evaluated)
2220, performance data Ti22230 of the device to be compared,
Input the dependent variable 2240) (step 320)
0).

The regression analysis data 220 input
Of 0, the number of function types 2210 and the number of types of devices to be evaluated and compared (two or more types) are the performance data type number 510 and the function type number 55 of the evaluation setting data 500, respectively.
Store in 0. And performance data Til2220, Ti2
2230 is used to perform the factor analysis process 2 described in the first embodiment.
Of 10, the processing of steps 211 to 214 is performed (step 3300).

Here, the explanatory variables used in the factor analysis processes 211 to 214 can be appropriately set according to the number of types of the above-described equipment to be evaluated and compared (that is, the number of types of performance data 510) (for example, the performance). Number of data types 51
If 0 is “2”, the three types of explanatory variables f1, f2, and f3 described in the first embodiment are set).

Among the calculated eigenvalues λ, the Gutman-Kaiser criterion and the scree test method (Fukumura Publishing, Psychological / Educational Data Analysis Method 10 Lecture-Basic Edition, p.
p.166-167), and select factors that affect user satisfaction with performance (see PC Statistical Analysis Handbook-II Multivariate Analysis, pp.198). The number of the selected factors is the number of factors 54 in the evaluation setting data 500.
It is stored in 0 (step 3400). Then, Example 1
Step 215 of the factor analysis process 210 described in
The processing of ˜217 is performed (step 3500).

Next, multiple regression analysis is performed using the factor score Z calculated in step 217 and the dependent variable 2240 of the regression analysis data 2200 (step 3600). Details of the multiple regression analysis are described in Personal Computer Statistical Analysis Handbook-II Multivariate Analysis, pp. 1-7. When the number of factors 540 is 2, in the multiple regression analysis processing 3600, the dependent variable Y2240 is used as the dependent variable, and the factor score Z (Zi1, Zi2) is used as the independent variable, and the dependent variable Y (as shown in Formula 7) is used. In order to express the above as a first-order polynomial of the factor score Z, the coefficients m1, m2 and m3 are determined by using the least squares method. Then, the coefficient is stored in the coefficient value 2120 of the evaluation formula setting data 2100, and the head pointer of the evaluation formula data 2100 is stored in the evaluation formula data pointer 570 of the evaluation setting data 500 (step 3700).

The evaluation formula can be changed by using the above-described derivation procedure.

Here, the performance evaluation of the equipment is performed for each kind of equipment to be evaluated (for example, Japanese word processor (WP),
A personal computer (PC), a workstation (WS), etc.) and each type of function to be evaluated (for example, character input, document editing, file operation, drawing, printing, etc.) are suitable. Therefore, the evaluation formula is derived for each target using the evaluation formula derivation process 3000 (for example, WP
It is conceivable to use an evaluation formula for the document editing function, an evaluation formula for the PC drawing function, etc.) and use an appropriate evaluation formula for each evaluation target.

As described above, in this embodiment, the evaluation formula for performing the performance evaluation in consideration of the user's satisfaction is derived (changed).
In such a case, the regression analysis data 2200 can be obtained by allowing the user to input only the data (that is, the regression analysis data 2200) necessary when deriving (changing) the evaluation formula.
The factor analysis process 210 is performed by using the performance data Til2220 and Ti22230, and the factor score Z obtained as a result is obtained.
Then, the multiple regression analysis process 3600 is performed using the dependent variable 2240 of the regression analysis data 2200 to derive (change) the performance evaluation formula considering the user's satisfaction.

Therefore, by mounting the means described in this embodiment in this system, the user can derive and change the evaluation formula for performing the performance evaluation in consideration of the user's satisfaction level as needed. You can

[0096]

According to the present invention, without using a method such as a questionnaire or a hearing survey, the response time of the device is used to automatically perform performance evaluation in consideration of user's satisfaction, and evaluate the evaluation result. Since the value can be quantitatively output, the product developer can easily and quickly perform the performance evaluation.

Further, by applying the evaluation method at the product development stage, it is possible to predict the opinion of the user in advance, and it is possible to extract the function having low user satisfaction before the product is commercialized. Therefore, the function can be rapidly improved, and a product satisfying the user can be put on the market.

[Brief description of drawings]

FIG. 1 is a diagram illustrating a hardware configuration of an information processing device.

FIG. 2 is a flowchart showing an outline of a performance measurement procedure executed by the information processing apparatus.

FIG. 3 is a diagram showing a configuration of evaluation setting data.

FIG. 4 is a diagram showing a performance data input display screen.

FIG. 5 is a diagram showing an evaluation result display screen.

FIG. 6 is a diagram showing an evaluation result registration screen.

FIG. 7 is a diagram showing a configuration of a performance evaluation list.

FIG. 8 is a diagram showing a structure of evaluation formula data used in Example 1;

FIG. 9 is a flowchart showing a performance data input process executed by the information processing apparatus.

FIG. 10 is a flowchart showing a performance evaluation process executed by the information processing device.

FIG. 11 is a flowchart showing a factor analysis process executed by the information processing device.

FIG. 12 is a flowchart showing an evaluation result output process executed by the information processing device.

FIG. 13 is a flowchart showing an evaluation result registration process executed by the information processing device.

FIG. 14 is a diagram showing a structure of evaluation formula data used in Example 2;

FIG. 15 is an evaluation formula used in performance evaluation.

[Explanation of symbols]

 1 ... Information processing apparatus, 100 ... Performance data input processing, 200 ... Performance evaluation processing, 210 ... Factor analysis processing, 220 ... Evaluation value calculation, 300 ... Evaluation result output processing, 400 ... Evaluation result registration processing, 500 ... Evaluation setting data , 600 ... Performance data input display screen, 700 ... Evaluation result display screen, 800 ... Evaluation result registration screen, 900 ... Performance evaluation list, 910 ... File management data, 920 ... Evaluation result data, 1000 ... Evaluation formula data (Example 1) ), 2000 ... Evaluation formula data (Example 2), 2100 ... Evaluation formula setting data, 2200 ... Regression analysis data, 3000 ... Evaluation formula derivation process.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Satoshi Kuwahara Inventor Teiji Kuwahara 292 Yoshida-cho, Totsuka-ku, Yokohama-shi, Kanagawa, Ltd. Hitachi, Ltd. Microelectronics Device Development Laboratory (72) Inventor Yoshihiko Nozoe Yoshida-cho, Totsuka-ku, Yokohama-shi, Kanagawa 292 Hitachi Ltd. Microelectronics Device Development Laboratory

Claims (5)

[Claims] 1. A computer having a function of performing statistical processing and various kinds of arithmetic processing, performs factor analysis processing using performance values for each function of information equipment, and calculates factor scores for each function as a result. , Multiple regression analysis processing is performed in combination with an evaluation value indicating the degree to which the user is satisfied with the performance (hereinafter, abbreviated as user satisfaction), and the regression equation obtained as a result is used to determine the user satisfaction. A performance evaluation formula that takes into consideration, a unit that stores the evaluation formula, a unit that inputs data, a unit that performs a factor analysis process using the data, a result of the factor analysis process, and a calculated factor score. A computer comprising: a means for calculating an evaluation value by substituting it into the stored evaluation expression; a means for outputting the evaluation value; and a means for registering the evaluation value. 2. The computer according to claim 1, wherein the data for performing the factor analysis process includes a performance value for each function of the information equipment. 3. The performance value according to claim 2, which is a response time from when a processing execution is instructed for each function of the information equipment until a processing result is obtained. The computer described in 1. 4. The explanatory variable set in the factor analysis process according to claim 1, using a performance value of a device for which performance evaluation is performed and a performance value of a device for comparison with the device, It includes the performance value of the evaluation target device itself, the difference between the performance value of the evaluation target device and the comparison target device, and the ratio of the performance value of the evaluation target device to the comparison target device. A calculator characterized by. 5. The performance value for each function of an information device (performance value for each function of a device to be evaluated or a device to be compared with the device) and an actual performance value for each function according to claim 1. A user's evaluation value, means for inputting evaluation formula derivation data that includes, factor analysis means for performing factor analysis processing by inputting performance values for each function of information equipment, and actual user's A multiple regression analysis method that performs multiple regression analysis processing using the evaluation value and the factor score calculated by the factor analysis processing, and a regression equation derived from the multiple regression analysis processing are used as performance evaluation equations that consider user satisfaction. A computer characterized by having, as means, a means for registering.