WO2007148867A1

WO2007148867A1 - Method of technology evaluation

Info

Publication number: WO2007148867A1
Application number: PCT/KR2007/000847
Authority: WO
Inventors: So Young Sohn; Tae Hee Moon
Original assignee: Industry-Academic Cooperation Foundation, Yonsei University
Priority date: 2006-06-20
Filing date: 2007-02-16
Publication date: 2007-12-27
Also published as: KR100776187B1

Abstract

The present invention relates to a technology evaluation method which reflects the characteristics of a target technology, the character of an enterprise possessing an associated technology, and the economic and environmental characteristics that the enterprise encounters for the purpose of being utilized in the technology finance field such as loans, credit guarantee, investment, etc., thereby enabling to predict the default probability by each year and the management achievement according to the technology evaluation rating. The inventive technology evaluation method is characterized by improving an evaluation model such that the distribution of the transcendently cognized total score of the evaluator can be adjusted while being contrastingly compared with the confidence interval of a posterior probability for an individual enterprise default obtained by a statistical modeling in order to take into consideration a phenomenon where an individual technology evaluation item score regarded as an input factor is re-assigned from the transcendently cognized total score based on the subjectivity of an evaluator, but is not evaluated independently, in the step of establishing a scoring model for calculating the technology commercialization risk score needed for obtaining the risk rating by using the past data.

Description

[DESCRIPTION]

[invention Title]

METHOD OF TECHNOLOGY EVALUATION

[Technical Field]

The present invention relates to a technology evaluation method which reflects the characteristics of a target

technology, the character of an enterprise possessing an associated technology, and the economic and environmental

characteristics that the enterprise encounters for the purpose of being utilized in the technology finance field such as

loans, credit guarantee, investment, etc., thereby enabling to

predict the default probability by each year and the management achievement according to the technology evaluation

rating. [Background Art]

In general, the technology evaluation can be roughly divided into two types, i.e., a technology performance assessment and a technology valuation. The technology

performance assessment can also be classified into a technology rating evaluation and a technology power evaluation. First, the technology rating evaluation is intended to evaluate the feasibility of technology commercialization on a percentage or rating basis. The technology power evaluation is intended to synthetically evaluate the overall technological capabilities of an enterprise, i.e., individual

technology level, technology manpower, intellectual property

right, research facility and so forth. In addition, the

technology valuation is aimed to convert the value of an individual technology into a monetary value so as to estimate the price of transaction technology, the collateral prices of

intellectual assets or the like.

In this case, a method of evaluating the technology

rating employs a scoring model, a profile model and a

checklist model. The scoring model is a method which scores respective evaluation items of the technical merit, and applies a weight value depending on a degree of importance between the evaluation items to thereby evaluate the

technology rating. The profile model is advantageous in that

evaluation elements are represented in a figure, and the merit and demerit of a corresponding technology can be rapidly

grasped. The checklist model is a method which sets up check points which need to be checked necessarily and establishes a cut-off criteria for the respective check points to thereby review whether all the check points satisfy a predetermined criteria value.

Such a conventional technology evaluation model is a model which determines respective evaluation elements and

allots each point (a degree of importance) to the respective evaluation elements to thereby evaluate the technology on a scoring and rating basis. This technology evaluation model

has a limitation in directly being utilized in the technology

finance field such as loans, investment, etc., since bankruptcy risk is not taken into consideration.

Moreover, the conventional technology evaluation method

has low evaluation reliability since it does not reflect the bankruptcy risk of enterprises according to economic and

environment or economic factor.

There has been recently proposed various technology evaluation methods to improve such problems. However, since technology evaluation items are set dependent on subjectivity

of an evaluator, distortion of the evaluation may occur and

the result of a technology scoring model established based on the distorted evaluation result may be also distorted.

Generally, detailed evaluation items used in a score evaluation method are composed of objective items that are evaluated equally irrespective of evaluators and subjective items having a high possibility of being fluctuated depending on evaluators. In terms of evaluation behavior of the evaluators for the objective items and the subjective items, lots of evaluators are liable to transcendentalIy cognize the synthesized scores for the items prior to evaluation of an

individual evaluation item. Thereafter, the scores conforming to the transcendentally cognized total scores are estimated

through the adjustment of individual scores for the subjective

detailed evaluation elements besides the objective evaluation

items with low degree of freedom of evaluation. Accordingly, in many cases, the individual evaluation item has a very low

significance whereas the synthesized scores have significance, which is the greatest merit of the score evaluation method.

Under this situation, if the relationship between the score of

the individual evaluation item and whether or not there occurs

an accident is grasped to obtain a weight value by each item, there is a possibility that erroneous scoring model will be induced. Therefore, there is a need for a technology

evaluation model which reflects the characteristics of classification of and the transcendentally cognized total

scores for the objective evaluation items and the subjective

evaluation items. In addition, the conventional technology evaluation method has a limitation in that a degree of importance of causes of the bankruptcy occurrence is not reflected by equally regarding all the causes of the accidents as a simple bankruptcy without taking into consideration various causes of the bankruptcy occurrence such as checking account bankruptcy,

interest delay, bad credit history, interest default and the like, and various types of accidents at the time of exhibiting the fluctuation trend of the enterprise bankruptcy over time.

Under this situation, Korean government announced "A comprehensive measure to enhance the competitiveness of the small Sc medium-sized enterprises" in July 2004. According to

this measure, there is a need for a credible technology rating and an improved authentication system for use in conformance

with various purposes such as the introduction of an improved

technology evaluation model for more advanced technology business feasibility evaluation, the construction of a rating establishment plan considering a bankruptcy rate according to technology commercialization, and so forth

[Disclosure] [Technical Problem]

Accordingly, the present invention has been made in an effort to solve the aforementioned problems occurring in the prior art, and it is an object of the present invention to provide a technology evaluation method which reflects the environmental risk and the technology commercialization risk according to economic environment, economic factor, etc. Another object of the present invention is to provide a technology evaluation method which re-constructs mutually

exclusive technology evaluation items based on a result of the analysis of a literature study and a significance of the past

technology evaluation index, and introduces a balance matrix

method to improve objectivity, neutrality and mutual

complementarity of the review by each item, thereby giving verified reliability to evaluation index and application

method . Still another object of the present invention is to provide a technology evaluation method which can exhibit the

fluctuation trend of the enterprise bankruptcy over time in

consideration of various causes of the enterprise bankruptcy occurrence . Yet another object of the present invention is to

provide a technology evaluation method which can analyze past similar cases of the technology evaluation of enterprises to

predict management achievements of new enterprises.

A further object of the present invention is to provide a technology evaluation method which provides credible technology evaluation authentication ratings for use in conformance with various purposes such as technology transfer, technology transaction, technology finance (loan and investment), etc., so as to promote the circumstances for activating the technology finance support for innovation-

driven small and medium-sized enterprises.

Another further object of the present invention is to

provide a technology evaluation method which can improve the

performance of an evaluation model by adjustably comparing the distribution of the transcendently cognized total score of an evaluator with the confidence interval of a good probability so as to reflect the transcendently cognized total score

rather than to independently estimate an individual technology evaluation item score considered to be an input factor in the

process of calculating technology commercialization risk scores needed for finding risk ratings.

Another yet further object of the present invention is to provide a technology evaluation method which differentially takes into consideration various causes of the enterprises bankruptcy through a competition risk model to improve a life data analysis model so as to enable to grasp a degree of a

default risk and to predict survival probability, so that the technology evaluation method can be applied to an enterprise rating evaluation as well as survival time and risk probability can predicted by reflection of the characteristics of bankruptcy enterprises to thereby contribute to figure out the cause of accidents. [Technical Solution]

To accomplish the above object, according to the present

invention, there is provided a technology evaluation method

comprising the steps of: obtaining a risk rating in which a probability-

calculated by a logit model is scored to produce scores and

ratings are assigned to the produced scores based on a

predetermined criterion so as to reflect a business bankruptcy risk of a technology to be evaluated; obtaining a technology rating in which ratings are

assigned to scores calculated by a weight scoring model based

on a predetermined criterion so as to reflect technical merit,

business potential and marketability of the to-be-evaluated technology; obtaining technology evaluation authentication rating in

which ratings are assigned by synthetically reviewing the risk

rating and the technology rating on a basis of a matrix whose

X axis indicates the technology rating and whose Y axis indicates the risk rating, wherein risk scores needed for obtaining the risk rating is calculated from a logit score and an environment score, wherein the logit score and the environment score are calculated by the following steps including: calculating a technology evaluation score factor; calculating an economic indicator factor; calculating an enterprise status factor;

determining a logit model input value based on basic

calculated values obtained through the above calculating

steps; inputting the calculated technology evaluation score

factor to a logit function 1 and calculating a good probability to thereby calculate the logit score; and inputting the calculated economic indicator factor and

enterprise status factor to a logit function 2 and calculating

a good probability to thereby calculate the environment score,

and wherein the steps of inputting the technology evaluation score factor to the logit function 1 and calculating the good probability comprises properly comparing the distribution of a

transcendently cognized total score of an evaluator with the confidence interval of a good probability obtained by the logit model so as to reflect the transcendently cognized total score, calculating an updated good probability, and calculating the logit score using the updated good probability.

Particularly, the distribution of the transcendently

corresponding to each evaluation item.

In addition, the technology evaluation method according

to the present invention further comprises a step of calculating a default probability by each year depending on ratings through a competition risk model. The steps of

calculating the default probability by each year comprises the steps of: calculating a linear combination of an input factor

and the coefficient; and calculating a survival probability

(1-default probability) depending on a given time. The step of calculating the survival probability comprises predicting and calculating the survival probability by using the

competition risk model which takes into consideration various characteristics and accident causes of an enterprise.

[Advantageous Effects] The technology evaluation method according to the

present invention has the following advantageous effect.

First, the inventive technology evaluation method

improves an evaluation model such that the distribution of the transcendently cognized total score of the evaluator can be adjusted while being contrastingly compared with the confidence interval of a posterior probability for an individual enterprise default obtained by a statistical modeling in order to take into consideration a phenomenon

where an individual evaluation item score is re-assigned from

the transcendently cognized total score based on the

subjectivity of an evaluator, but is not evaluated independently, in the step of establishing a scoring model for calculating the technology commercialization risk score needed for obtaining the risk rating. By doing so, less adjustment

is made in the probability interval whose prediction accuracy

is high whereas much adjustment is systematically reflected in the probability interval whose prediction accuracy is low of the confidence interval of the posterior probability for the

enterprise default, so that the technology evaluation can be performed which systematically reflects the improvement in the

performance of the technology evaluation model and the

cognition scoring by a know-how of the evaluator.

Second, a life data analysis model is improved so that a degree of default risk over time can be grasped by reflecting

the type of various causes (for example, interest delay, checking account bankruptcy, bad credit history, etc.) of enterprise bankruptcy associated with technology finance through the competition risk model. Thus, the technology finance organization aimed at loans, credit guarantee, investment, etc., can apply the inventive technology evaluation method to an assessment of the feasibility of

business potential for the technology as well as can predict

the survival time period and the risk probability of an

enterprise in bankruptcy by reflecting the characteristic of the enterprise, thereby contributing to diagnosis of the cause

of the enterprise bankruptcy. [Description of Drawings]

The above and other objects, features and advantages of

the present invention will be apparent from the following

detailed description of the preferred embodiments of the invention in conjunction with the accompanying drawings, in which:

FIG. 1 is a diagrammatic view illustrating a concept of a technology evaluation method according to a preferred

embodiment of the present invention;

FIG. 2 is a flowchart illustrating a process of calculating an evaluation rating by each technology evaluation item according to the present invention;

FIG. 3 is a flowchart illustrating a process of calculating risk scores (logit model) needed for finding a risk rating according to the present invention;

FIG. 4 is a flowchart illustrating a process of predicting the management achievement according to the present invention;

FIG. 5 is an exemplary screen illustrating a diagram

section and a commentary section of a KTCP matrix;

FIG. 6 is a flowchart illustrating a process of constructing a KTCP matrix;

FIG. 7 is a KTCP matrix diagram showing a technology evaluation rating;

FIG. 8 is a KTCP matrix diagram showing a management

achievement prediction;

FIG. 9 is a KTCP matrix diagram showing technical merit versus business potential;

FIG. 10 is a KTCP matrix diagram showing technical merit versus marketability;

FIG. 11 is a KTCP matrix diagram showing marketability

versus business potential;

FIG. 12 is a KTCP matrix diagram showing financial stability versus technology's business potential;

FIG. 13 shows a receipt management screen of a technology evaluation receipt step; FIG. 14 shows an input screen of receipt details in a technology evaluation receipt step;

FIG. 15 shows an input screen of an enterprise profile;

FIG. 16 shows an input screen of the personal information status of a representative;

FIG. 17 shows an input screen of financial statements;

FIG. 18 shows a management screen of an economic and

environmental indicator data;

FIG. 19 shows a creation screen of an evaluation form

cover;

FIG. 20 shows a screen on which to draw up an evaluation table through the input of 45 data;

FIG. 21 is an evaluation item rating assignment screen

of an exemplary measurement item 1;

FIG. 22 is an evaluation item rating assignment screen of an exemplary measurement item 2 ;

FIG. 23 is an evaluation item rating assignment screen

of an exemplary check item; FIG. 24 is an evaluation item rating assignment screen

of an exemplary evaluator's evaluation item;

FIG. 25 is a screen showing a balance matrix;

FIG. 26 is a screen showing a score calculation result;

FIG. 27 is a screen showing the calculation results of technology evaluation ratings, achievement analysis, survival probability (=l-default probability) in an evaluation result display step;

FIG. 28 is a graph showing the comparison of business bankruptcy prediction powers between a conventional technology-

evaluation method in which technology risk ratings are not

reflected and a technology evaluation method according to the

present invention; and FIG. 29 is an ROC graph showing the comparison of business bankruptcy prediction powers between a conventional technology evaluation method in which technology risk ratings

are not reflected and a technology evaluation method based on a cognition scoring model according to the present invention. [Best Mode]

Now, a preferred embodiment of a technology evaluation

method according to the present invention will be described hereinafter in detail with reference to the accompanying

drawings . FIG. 1 is a diagrammatic view illustrating a concept of

a technology evaluation method according to a preferred

embodiment of the present invention.

Referring to FIG. 1, the technology evaluation method according to the present invention comprises a step S123 of finding a risk rating, a step S125 of finding a technology rating, a step S129 of finding a technology evaluation authentication rating, a step S133 of finding default probabilities by year depending on the ratings, and a step S135 of predicting the management achievement, each of which

will be described below.

The evaluation rating of the technology evaluation

method according to the present invention is roughly divided into a risk rating (S123), a technology rating (S125) and a

technology evaluation (authentication) rating (S129) .

The risk rating (S123) is a technology-based enterprise

risk rating which statistically analyzes commercialization conditions, technical merits and associated matters of the

technology of which commercialization will be made or is in

progress so as to review a business bankruptcy possibility of

enterprises based on a corresponding technology and simultaneously synthesize economic and environmental factors, etc., that can affect the business propulsion of the

enterprises. The risk rating is obtained by scoring a good or bad probability calculated by a logit model to produce scores and assigning ratings to the produced scores based on a

predetermined criterion so as to reflect a business bankruptcy risk of an enterprise (technology) to be evaluated, and is represented by ten ratings ranging from "aaa" to "d" .

The technology rating (S125) is a technology-based raging which synthetically reviews technical merit, business potential and marketability of the technology of which commercialization will be made or is in progress so as to

calculate business success possibility and technological innovation capability of enterprises based on a corresponding

technology. The technology rating is obtained by assigning

ratings to scores calculated by a weight scoring model based

on a predetermined criterion so as to reflect technical merit, business potential and marketability of the enterprise (technology) to be evaluated, and is represented by ten ratings ranging from "Vl" to "VlO" .

The technology evaluation (authentication) rating (S129) is a rating which is assigned by synthetically reviewing the

risk rating and the technology rating on a basis of a matrix

composed of an X-axis and an Y-axis. The technology evaluation (authentication) rating is represented by ten ratings ranging from "AAA" to "D" .

Such an evaluation rating display system is shown in

Table 1 below.

[Table 1)

Here, the risk rating (S123) and the technology rating

(S125) are ratings which are assigned based on a predetermined criteria depending on risk scores sand technology scores calculated by different models, and the technology evaluation

(authentication) rating (S129) is a rating which is assigned

by synthetically reviewing the risk rating and the technology rating on a basis of a matrix whose Y axis indicates the risk

rating and whose X axis indicates the technology rating. A

result of the technology evaluation (authentication) rating is

shown in Table 2 below.

[Table 2]

Now, a method of finding the risk rating, the technology- rating and the technology evaluation authentication rating based on the technology evaluation method according to the

present invention will be described in detail hereinafter.

Technology evaluation items used to find the risk rating, the technology rating and the technology evaluation authentication rating in the technology evaluation method are shown in Table 3 below. The technology evaluation items are composed of a hierarchical structure including a large item, a

middle item, a small item and a detailed review item. In this case, a score (rating) by each technology evaluation item (small item) is calculated on the basis of an evaluation

result of the detailed review item, and the score (rating) of

this technology evaluation item (small item) is utilized as input factors for calculation of the risk rating and the

technology rating.

FIG. 2 is a flowchart illustrating a process of calculating an evaluation rating by each technology evaluation item according to the present invention.

As shown in FIG. 2, an evaluator logs in to a technology evaluation table (S201) , and inputs the evaluation contents of

detailed review items by each technology evaluation item (small item) (S203) .

The input evaluation contents are divided into three categories depending on an evaluation mode as follows: measurement evaluation item, check evaluation item and evaluator's evaluation item.

The measurement evaluation item is an item which is

evaluated based on values calculated automatically by means of

a system by utilizing an enterprise-related data that is input separately such as enterprise profile data, enterprise

financial data, etc. Since such a measurement evaluation item is automatically calculated by the system, it is very important to check whether data input is omitted.

The measurement evaluation item is roughly classified

into an evaluation item based on the size (level) of a corresponding calculated value itself, and an evaluation item

based on a level to the average value (average ratio by each business line) of the same industry. The calculation method by each item is shown in Table 4 below. measurement evaluation item (reviews)

Reviews Input value Output value Calculation method

The check evaluation item is automatically evaluated by

the number of checked items by allowing an evaluator to check the check items suggested as evaluation criteria.

The check evaluation item status is shown in Table 5

below.

[Table 5]

( Iheck evaluation item (reviews)

Reviews Input values Output values Calculation method technology Check The number of the evaluator management checked cases counts the strategy number of items checked on an evaluation pop-up screen to perform evaluation by total number of checked items . technical the same as the same as the same as manpower above above above management crisis the same as the same as the same as management above above above capability management the same as the same as the same as commitment and above above above business acumen relationship the same as the same as the same as with above above above representative and teamwork

technology the same as the same as the same as spillover above above above effect on the enterprise inside and outside competitiveness the same as the same as the same as situation above above above within the same industry- market the same as the same as the same as accessibility above above above recognition the same as the same as the same as level above above above comparison the same as the same as the same as superiority to above above above alternative product easiness of the same as the same as the same as materials and above above above parts procurement validity of the same as the same as the same as sales plan above above above

The object of evaluator's evaluation item is an evaluation item other than the measurement and check evaluation items. Each evaluator' evaluation item is given an

arbitrary rating by an evaluator so as to be evaluated based on an evaluation criterion suggested on an evaluation (pop-up) screen.

The evaluator's evaluation method by each evaluation item is summarized in Table 3 as described above.

An evaluator inputs the contents of the detailed review items by each technology evaluation item (small item) based on the above evaluation method (S203) , and then determines whether or not the input of any detailed review item is

omitted (S205) . If there exists any omitted detailed review

item, the evaluator inputs the omitted review items (S207) .

When all the detailed review items are input, the score (rating) by each technology evaluation item (small item) is automatically calculated depending on the number of detailed

review items by each technology evaluation item (small item)

based on a 5x5 balance matrix or a 5x5x5 balance matrix (S209) . An example of calculation of the score (rating) by a technology evaluation item (small item) based on the 5x5

balance matrix is shown in Table 6 below.

[Table 6]

O

£■

"o A

B

C

D

E

E D C B A ratings scores counts probability application and extendability of technology Also, an example of calculation of the score (rating) by a technology evaluation item (small item) based on the 5x5x5

balance matrix is shown in Table 7 below.

[Table 7] technology high

technology knowledge level

FIG. 3 is a flowchart illustrating a process of calculating risk scores (logit model) needed for finding a risk rating according to the present invention. A risk prediction model is established by applying the technology evaluation data evaluated on the technology evaluation table

to a logit model .

A process of calculating the logit model needed for

finding the risk rating first includes a step of checking if

the input of the evaluation contents of the detailed review

items by each technology evaluation item (small item) is omitted (S301) . In this step S301, whether or not the scoring

of the evaluation items is omitted and the input of other input factors is omitted is checked automatically. As a

result of the checking, if the input of any evaluation score

is omitted, the item is complemented (S303) .

Subsequently, a technology evaluation score factor based

on a functional expression is calculated by utilizing the

evaluated and input scores (ratings) by each technology evaluation item (small item) .

The factor calculation expression is shown in Table 8

below.

[Table 8]

[Table 9]

Since the enterprise state factor is a factor whose input error or input omission exists in the process of the regular survey and data management DB input, it is configured

to pop-up an enterprise status factor list on a screen so as to allow an evaluator to re-confirm the enterprise status

factor prior to calculation of a final score.

Next, the basic calculated values and a logit model input value are determined and stored (S313) .

Subsequently, the calculated technology evaluation score factor is input to a logit model to obtain a logit score 1

(S315) and calculating a good probability to calculate the logit score (S319) .

The item name and factor name for calculating the logit score are shown in Table 11 below.

[Table ll]

Here, the present invention mainly features that the

transcendently cognized total score estimated from experiences

and know-how of the evaluator is reflected in the course of

calculating the good probability using the logit model .

The detailed evaluation item generally used in the score evaluation method is roughly divided into an objective item

which is equally evaluated irrespective of the subjective thinking of an evaluator and a subjective item whose

evaluation is highly likely to be fluctuated depending on an

evaluator. In regard to the evaluation behavior of the evaluator relative to the detailed evaluation item, in many

cases, an evaluator typically cognizes the synthesized scores

transcendently prior to the evaluation of respective items. Thereafter, the evaluator

The score is calculated which conforms to the transcendently cognized total score through an individual score adjustment for subjective detailed evaluation elements besides the objective evaluation items which are low in a degree of freedom of evaluation. Thus, in many cases, an

individual evaluation item has a low significance whereas the synthesized score has significance. As a result, such a

characteristic is the greatest merit of the score evaluation method. Under such circumstances, if the relationship between

the individual item score and the accident is grasped to

obtain a weight value by each item, there is a possibility of inducing an erroneous scoring model. Therefore, there is a

need for an evaluation model which reflects the characteristics of classification of the objective evaluation item and the subjective evaluation item and the transcendently

cognized total score.

To this end, in the present invention, analysis of the

relativity with non-accident is carried out using only the

objective evaluation item which is similarly evaluated depending on evaluators based on a logistic regression analysis. That is, this model is a model in which the

transcendental influence of the evaluator is excluded.

First, the logistic regression analysis equation is written as follows:

P= P(y = \\X₀)=F(X₀'β) ₍₁₎

where, P denotes a default probability,

(4)

2 where, ⁽l-a/2) ^_enotes a standard normal distribution,

and

.

Further, X°'^β^LB and X°'β"^UB denote a lower limit and an upper

limit of the confidence interval of Xo'β" .

In terms of the characteristic of such a logistic

regression model, the interval having a posterior probability

of about 0 or 1 is an interval whose discriminatory power is

high for an evaluation result by the model, and whose confidence interval is very narrow. On the contrary, the

interval having a posterior probability of about 0.5 is an

interval whose discriminatory power is high of the evaluation result, and whose confidence interval is very wide. That is,

in consideration of such characteristic, the transcendentally

cognized total score of the evaluator is reflected by the following equation in such a fashion as to be reflected slightly in the interval whose discriminatory power is high and reflected greatly in the interval whose discriminatory power is low in the above analyzed model, so that it is possible to establish a technology evaluation model which can

properly combine a mechanical model and a human cognitive

cognized total score.

The updated good probability (score) is calculated by

1 - P (S317) , and the logit score is calculated by an

equation of 'logit score = updated good probability x 100'

(S319) .

Conclusively, such adjustment method is performed in such a fashion that the distribution of the transcendently cognized total score of the evaluator is adjusted by comparing

the distribution of the transcendently cognized total score

with a confidence interval of a posterior default probability

(or good probability) , so that less adjustment is made in the probability interval whose prediction accuracy is high whereas much adjustment is systematically reflected in the probability interval whose prediction accuracy is low.

Next, the item name and factor name for calculating the environment score are shown in Table 12 below. The economic indicator factor and the enterprise status factor is input to the logit model to calculate a logit score 2 (S321) , and then

a good probability is calculated (S323) to obtain the environment score (S325) .

[Table 12]

If the logit score and the environment score are obtained, then the risk score is obtained by an equation of

'risk scores = logit score x + environment score x

Ql - W) , (S327) . in this case, ^ is a weight value for

weight combination of the logit score and the environment score .

If the risk score is obtained, then as shown in FIG. 1,

then the risk rating is calculated (S123) .

A subsequent step is a step of obtaining the technology rating based on the weight scoring model (S125) . There is

illustrated a method of obtaining the technology score based on the technology evaluation item (small item) .

[Table 13]

A weight by each evaluation item reflects the characteristic by each business line, and differentially

adopts a weight by each business line.

When the technology score is determined by the weight scoring model, the technology rating is assigned to the technology score (S125) based on an evaluation criteria shown in Table 14 below.

[Table 14]

(where, the technology rating is represented by upper- case letters "Vl to VlO") When the risk rating (S123) is calculated by the logit model and technology rating (S125) is calculated by the weight

scoring model, a step of finally assigning the technology evaluation authentication rating (S129) will be described by a

matrix evaluation criterion on basis of the risk rating and

the technology rating.

The technology evaluation authentication rating is calculated by the rating evaluation according to the matrix of

Table 2 as described above. Next, a step of calculating the default probability (=1-

survival probability) by each year depending on the rating by

the life data analysis model (survival analysis model) will be described (S133) . This step is a step where an evaluation result of the technology evaluation table is used as an input

factor to calculate the default probability (=1- survival

probability) by each year depending on the technology evaluation authentication rating.

The life (survival) data is composed of the time period ranging from a predetermined start time point to an event occurrence time point, for example, from a time point when a credit guarantee (loan) is made depending on the technology evaluation and evaluation result to a time point when an event of enterprise bankruptcy (failure in technology commercialization) occurs, and this time period is called a survival period.

The life data analysis (survival analysis) provides a

fluctuation trend of a business bankruptcy rate depending on the technology evaluation rating by analyzing the survival

period to find a survival probability and a risk rate since a t time point of the technology evaluation and credit guarantee (loan) enterprise so as to measure the expected life of an to-

be-evaluated enterprise. However, a general life data analysis has a

shortcoming of not suggesting a degree of risk considering the

correlation between various causes (interest delay, checking

account bankruptcy, bad credit history) of the enterprises bankruptcy. The present invention is characterized in that a

competition risk model is established based on the Weibull

distribution in the step of calculating the survival probability, in which time a scale parameter of the Weibull distribution is composed of a regression model in view of an input factor reflecting the characteristics of borrower or the characteristics of obligor, a shape parameter of the Weibull

distribution by each competition risk cause is simulated differentially. T ^{■ ■}

The following equation (6) shows the time ^ij' spent for an enterprise i to encounter a bankruptcy or default occurs

due to j-th cause based on the assumption of the Weibull distribution.

Here, the present invention mainly features that the transcendently cognized total score estimated from experiences and know-how of the evaluator is reflected in the course of calculating the good probability using the logit model. where, i(i=l,...,n) denotes an individual enterprise,

j^th (j =1, ... , k) denotes an accident cause by which a default

CL 7 ^■ occurs, and ^ij and ^J denote a scale parameter and a shape parameter, respectively, ^u^ ' and ^iJ of such an Weibull distribution are a probability density function and an

accumulated probability density function, respectively. The

survival function is obtained by the following equations (7)

to (9) using the probability density function and the accumulated probability density function:

^^.(0 =l-exp[-(α,0^r;] ₍₈₎

where, a probability that the enterprise I will be survived with respect to the cause j, i.e., ^ quantile life is written by the following equation (10) :

(10) γ . where, parameters ' ■> of the Weibull distribution is

different from each other by each enterprise accident cause,

I/a.. exptX.β,) and if other parameter ¹J is processed with ^J to

reflect the property variable ( ) of the enterprise i to achieve regression simulation, this regression simulation can

be represented by the following equation (11) :

(11) •

Using the above equation (8) , a survival probability s^lO) that no accident will occur with respect to any accident

cause, and a probability that any accident will occur due to a certain cause can be obtained by the following

equations (12) and (13) :

The survival probability of the enterprise i can be obtained from the above equation (12), and the default probability of the enterprise i can be obtained from the above equation (13) .

By performing this method, a survival probability

for an individual accident cause as well as a survival

probability considering all the accident causes can be predicted, and hence can be applied under various purposes.

The input factor list needed for calculating the survival probability of the life data analysis model is shown in Table 15 below.

[Table 15]

Table 16 as shown below is an example of shape parameters and time set values for calculating a survival probability by each year. In this case, the time is set such that one year consists of 12 months on a basis of a month unit.

[Table 16]

The default probability (1-survival probability) of a

time point t given by such factors is calculated by first calculating a linear combination of the input factor and the

coefficient and second calculating the default probability (=l-survival probability) of the given time point t.

An example of an estimated result of the default

probability (=l-survival probability) of the given time point

t is shown in FIG. 27.

Next, a step (S135) of predicting the management achievement based on the case-based reasoning (S121) will be

described hereinafter.

The case-based reasoning (CBR) is the process of searching for a past case similar to a new event and solving

new problems using similarity to the solutions to problems of

the similar past case. By application of this analysis method, the past technology evaluation case similar to that of the

enterprise (technology) being evaluated currently is extracted and the management achievement of an enterprise having the

similar technology evaluation case is analyzed to thereby extract the management achievement of a corresponding

enterprise . FIG. 4 is a flowchart illustrating a process of

predicting the management achievement according to the present

invention.

The management achievement predicting process will be described step by step hereinafter with reference to FIG. 4.

First, a score by each technology evaluation index is

calculated (S401) . In this step, the score calculated by the technology evaluation table is utilized as it is.

Subsequently, a Euclidean distance is measured (S403) .

The Euclidean distance is an index indicative of the criteria of similarity, and is obtained by a square root of a sum of squares of the distance between a reference score of the technology evaluation small item of an to-be-evaluated enterprise and a score of a to-be-compared enterprise (past scoring case) . The Euclidean distance measuring equation is

as follows:

16

Weighted Euclidean distance = J∑ Σ((P_n ~P_nY *WJ

«=1

: score of a to-be-compared enterprise

(past scoring case) p ⁿ : score a to-be-evaluated enterprise

Wⁿ : weight

n = 1, 2, ... , 16

^ Calculation of the Euclidean distance for each to-be-

compared enterprise (past scoring case)

Next, the cases of 10 enterprises having the minimum Euclidean distance, i.e., the highest similarity are extracted are extracted (S405) through the measurement of the Euclidean

distance, and the management achievement of growth potential,

profitability, turn-over, etc., of 10 enterprises are extracted (S407) .

The achievement evaluation result based on the case- based reasoning for the to-be-evaluated enterprise is accumulated in that of the to-be-compared enterprise (past scoring case) so as to be utilized as the to-be-compared enterprise. In the foregoing, the risk rating, the technology rating, the technology evaluation authentication rating, the

default probability by each year, and the management achievement prediction have been described.

In case where the technology evaluation result according

to the technology evaluation method of the present invention as described above is analyzed from various points of view including risk rating, technical merit, business potential,

etc., and is displayed on a two-dimensional diagram

(hereinafter, referred to as ^ΛKTCP matrix')/ its utility can increase. Thus, the drawing method and utility of such a KTCP

matrix will be descried hereinafter.

FIG. 5 is an exemplary screen illustrating a diagram section and a commentary section of a KTCP matrix. As shown in FIG. 5, the KTCP matrix is a technology evaluation an consulting matrix which is composed of a diagram section and a commentary section so as to

represent the characteristic of the enterprise in the form of

a diagram having the technical merit and the business potential, the technology's business potential and the risk rating factor, etc., indicated in an X-axis and an Y-axis.

The diagram section is configured such that a technical merit and a business potential, and a risk viewpoint evaluation result of the to-be-evaluated enterprise is sorted into certain ratings and then is represented on the matrix.

The commentary section is configured such that the

matrix is combined and constructed from various points of view,

and then the explanation for each matrix is automatically output .

The technical merit used as an index of the KTCP is evaluated by synthesizing technology development propulsion

capacity, investment and infrastructure for technology development, technology innovation, degree of technology completion, technology extendability, etc., and the business

potential is evaluated by synthesizing market status and

competition of a retained technology (product) , commercialization capability, marketability, etc. FIG. 6 is a flowchart illustrating a process of constructing a KTCP matrix.

As shown in FIG. 6, the process of constructing the KTCP

matrix comprises a step (S601) of selecting a factor, a step

(S603) of calculating a risk rating, a technology rating, and a financial stability rating, a step (S605) of calculating the scores of a technical merit, a business potential, and a technology's business potential, a step (S607) of establishing the interval of scores by each factor, a step (S609) of drawing up a matrix table, and a step (S611) of constructing the diagram section and the commentary section.

In the factor selecting step (S601) , each factor of the technology evaluation risk rating, the technology rating and

the technical merit factor is selected. The technology

evaluation risk rating employs a technology-based risk rating

calculated by the technology evaluation model, and the technology rating employs a score or rating calculated by the

scoring method with respect to the entire evaluation items of the technology evaluation index as described above.

As shown in Table 17, the factor of the technical merit

is composed of technical merit evaluation items of the

technology evaluation index.

[Table 17]

As shown in Table 18, the factor of the marketability is composed of marketability evaluation items of the technology evaluation index.

[Table 18]

Table 19 shown below is the factor of the business

potential 1.

[Table 19]

Table 20 shown below is the factor of the business potential 2.

[Table 2θ]

As shown in Table 21, the factor of the technology's business potential is composed of technical merit and business potential evaluation items of the technology evaluation index.

[Table 21]

Large item Middle item Small item

The factor of the financial stability employs a financial rating utilizing a financial model of an enterprise credit evaluation system.

If the factor selecting step is completed, then the risk

rating, the technology rating, the financial stability score

(rating) are calculated (S603) .

The risk rating is a technology-based risk rating calculated by the technology evaluation model as described

above .

The technology rating is a rating calculated by the

technology evaluation model as described above.

The financial stability rating is a financial rating calculated by the financial model of the enterprise credit evaluation system.

Subsequently, the scores of the technical merit, the

business potential and the technology's business potential are calculated (S605) .

The score by each field (factor) is converted and

calculated relative to a full 100 score according to the following calculation equation by the scoring model based on the scoring method.

■ Calculation equation score by each factor ∑ (evaluated score of each techn ology evaluation item x weight) full score by ^ (maximum score of each techn ology evaluation item(5) x weight) each factor 67 Full score by each factor

If the score is calculated, an interval by each factor is established (S607) such that a score (rating) by each factor is first sorted into six intervals so as to construct

the diagram section.

Risk rating .VS. technology rating Matrix constitute 10 x 10

Financial stability VS technical merit/business potential Matrix (6 x 6)

_Q

CO

^■a

4) O υ

<

⁽β

^■a

O

C CO

C

technical merit /business potential(a) acceptable

Finally, the diagram section and the commentary section of

the matrix are constructed (S611) .

FIG. 7 is a KTCP matrix diagram showing a technology

evaluation rating. In FIG. 7, a matrix is drawn up whose X

axis indicates the technology rating and whose Y axis indicates the risk rating. The technology rating means that the feasibility of technology commercialization is evaluated as a rating based on the technology power, the technology development status and the commercialization capability in association with a to-be-evaluated technology, and the risk rating means that the default possibility of the future

business predicted in the course of a technology business propulsion in view of the environment inside and outside a to- be-evaluated enterprise is evaluated as a rating based on the risk evaluation model.

FIG. 8 is a KTCP matrix diagram showing a management

achievement prediction.

In FIG. 8, in view of the fact that a technology-based

enterprise, particularly a small and medium-sized enterprise has much more difficulty predicting the future management achievement as compared to a conventional traditional type

enterprise, various management achievement cases of the technology-based enterprise are analyzed and the next three

year-term management achievement of the technology-based enterprise is predicted on a basis of such analysis. The

management achievement is calculated based on four financial indices including profitability, growth potential, stability and turn-over. The fact that the rating of the management achievement is higher means that better achievement is predicted.

FIG. 9 is a KTCP matrix diagram showing the technical merit versus the business potential. The KTCP matrix diagram of FIG. 9 is a diagram which contrastingly represents the technology content, the technology level, etc., of a to-be- evaluated technology, and the commercialization capability,

marketing situation, future profitability, etc., of the technology on the matrix.

FIG. 10 is a KTCP matrix diagram showing the technical merit versus the marketability. The KTCP matrix diagram of FIG. 10 is a diagram which contrastingly represents the

technology content, the technology level, etc., of a to-be-

evaluated technology, and the market size, growth prospect,

competition situation, etc., of the to-be-evaluated technology on the matrix.

FIG. 11 is a KTCP matrix diagram showing the

marketability versus the business potential. The KTCP matrix diagram of FIG. 11 is a diagram which contrastingly represents

the market size, growth prospect, competition situation, etc., of the to-be-evaluated technology, and the commercialization capability, marketing situation, future profitability, etc.,

of the technology on the matrix.

FIG. 12 is a KTCP matrix diagram showing the financial stability versus the technology's business potential. The KTCP matrix diagram of FIG. 12 is a diagram which contrastingly represents future growth prospect (technology's business potential) such as the technical merit, business potential, profitability, etc., of the to-be-evaluated technology, and

short-term financial stability of the evaluation time point

for the enterprise on the matrix.

At next steps, there is shown each screen on a computer which actually applies the technology evaluation method according to the present invention as described above. Each screen displayed depending on an evaluation order of the

present invention will be described hereinafter.

FIG. 13 shows a receipt management screen of a technology evaluation receipt step, and FIG. 14 shows an input screen of receipt details in the technology evaluation receipt step.

FIGs. 15 to 18 show an input screen of an enterprise data, wherein FIG. 15 is an input screen of an enterprise

profile (business line, business history, registration, external audit, venture, etc.), FIG. 16 shows an input screen of the personal information status of a representative, FIG.

17 shows an input screen of financial statements, and FIG. 18

shows a management screen of an economic and environmental indicator data.

FIGs. 19 to 26 show input steps of the evaluation data, wherein FIG. 19 shows a creation screen of an evaluation form cover, FIG. 20 shows a screen on which to draw up an evaluation table through the input of 45 data, FIG. 21 is an

evaluation item rating assignment screen of an exemplary

measurement item 1, FIG. 22 is an evaluation item rating assignment screen of an exemplary measurement item 2, FIG. 23 is an evaluation item rating assignment screen of an exemplary

check item, FIG. 24 is an evaluation item rating assignment

screen of an exemplary evaluator's evaluation item, and FIG.

25 is a screen showing a balance matrix, FIG. 26 is a screen

showing a score calculation result. FIG. 27 is a screen showing the calculation results of technology evaluation ratings, achievement analysis, survival probability (=l-default probability) in an evaluation result

display step.

Also, FIG. 28 is a graph showing the comparison of business bankruptcy prediction powers between a conventional technology evaluation method in which technology risk ratings are not reflected and a technology evaluation method according

to the present invention, and FIG. 29 is an ROC graph showing the comparison of business bankruptcy prediction powers between a conventional technology evaluation method in which technology risk ratings are not reflected and a technology

evaluation method based on a cognition scoring model according to the present invention. According to FIG. 28, although a conventional technology

evaluation method in which technology risk ratings are not

reflected must exhibit a higher bankruptcy (default) rate in a lower rating, it does not exhibit it. According to FIG. 29, the conventional technology evaluation method also exhibits a lower bankruptcy (default) rate in the comparison of the

performance evaluation by a lower area of the ROC graph. On

the other hand, the technology evaluation method according to the present invention sequentially exhibits lower bankruptcy

(default) rates in higher ratings, and shows an improved result in the comparison of the performance evaluation by a lower area of the ROC graph as compared to the conventional

technology evaluation method.

As mentioned above, the technology evaluation method

according to a preferred embodiment of the present invention is embodied according to the program necessary for implementing the present invention through the data stored in

a database of a computer,

[industrial Applicability] As described above, the technology evaluation method according to the present invention has the following advantageous effect.

First, the inventive technology evaluation method improves an evaluation model such that the distribution of the transcendently cognized total score of the evaluator can be

adjusted while being contrastingly compared with the

confidence interval of a posterior probability for an individual enterprise default obtained by a statistical modeling in order to take into consideration a phenomenon

where an individual evaluation item score is re-assigned from the transcendently cognized total score based on the

subjectivity of an evaluator, but is not evaluated

independently, in the step of establishing a scoring model for calculating the technology commercialization risk score needed for obtaining the risk rating. By doing so, less adjustment

is made in the probability interval whose prediction accuracy is high whereas much adjustment is systematically reflected in

the probability interval whose prediction accuracy is low of the confidence interval of the posterior probability for the enterprise default, so that the technology evaluation can be performed which systematically reflects the improvement in the

performance of the technology evaluation model and the cognition scoring by a know-how of the evaluator.

the type of various causes (for example, interest delay, checking account bankruptcy, bad credit history, etc.) of enterprise bankruptcy associated with technology finance

through the competition risk model. Thus, the technology

finance organization aimed at loans, credit guarantee,

investment, etc., can apply the inventive technology evaluation method to an assessment of the feasibility of business potential for the technology as well as can predict the survival time period and the risk probability of an

enterprise in bankruptcy by reflecting the characteristic of

the enterprise, thereby contributing to diagnosis of the cause of the enterprise bankruptcy.

While the invention has been described in connection with what is presently considered to be practical exemplary

embodiments, it is to be understood that the invention is not

limited to the disclosed embodiments, but, on the contrary, is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the

appended claims.

Claims

[CLAIMS]

[Claim l]

A technology evaluation method comprising the steps:

obtaining a risk rating in which a probability

calculated by a logit model is scored to produce scores and ratings are assigned to the produced scores based on a

assigned to scores calculated by a weight scoring model based

on a predetermined criterion so as to reflect technical merit, business potential and marketability of the to-be-evaluated technology;

obtaining technology evaluation authentication rating in

which ratings are assigned by synthetically reviewing the risk rating and the technology rating on a basis of a matrix whose

X axis indicates the technology rating and whose Y axis indicates the risk rating,

wherein risk scores needed for obtaining the risk rating is calculated from a logit score and an environment score, wherein the logit score and the environment score are calculated by the following steps including: calculating a technology evaluation score factor; calculating an economic indicator factor;

calculating an enterprise status factor;

determining a logit model input value based on basic calculated values obtained through the above calculating

steps; inputting the calculated technology evaluation score factor to a logit function 1 and calculating a good probability to thereby calculate the logit score; and inputting the calculated economic indicator factor and enterprise status factor to a logit function 2 and calculating

a good probability to thereby calculate the environment score,

and wherein the steps of inputting the technology evaluation score factor to the logit function 1 and calculating the good

probability comprises properly comparing the distribution of a

transcendently cognized total score of an evaluator with the

confidence interval of a good probability obtained by the

logit model so as to reflect the transcendently cognized total score, calculating an updated good probability, and calculating the logit score using the updated good probability.

[Claim 2]

The technology evaluation method of claim 1, wherein the distribution of the transcendently cognized total score of the

corresponding to each evaluation item.

[Claim 3]

The technology evaluation method of claim 1 further comprising a step of calculating a default probability by each

year depending on ratings through a competition risk model, wherein the steps of calculating the default probability

by each year comprises the steps of: calculating a linear

combination of an input factor and the coefficient; and

calculating a survival probability (1-default probability) depending on a given time, and wherein the step of calculating the survival probability

comprises predicting and calculating the survival probability by using the competition risk model which takes into

consideration various characteristics and accident causes of

an enterprise.