Evaluation Model Learning & Resource > About SMART3 > Evaluation Model

Evaluation Model

Evaluation Model


Identifying and scoring patent indicators for patent evaluation

Evaluation Indicators : Strength of Patent Rights, Quality of Technology, Usability

Strength of Patent Rights Quality of Technology Usability
The degree to which you can
maintain exclusive status in a patent
dispute with a third party
To match or
lead technology trends
The degree and
availability of business
35% 35% 30%


Determine assessment elements using factual information that can be extracted
for patent evaluation

Evaluation Factor condition : Objectivity, Quantitativeness and Completeness

Objectivity Quantitativeness Completeness
A completely independent nature
from personal supervision
Quantify statistical
observations, including
mathematical meanings
The information that
any patents eligible for

KR Patent Evaluation Model

Structural equation evaluation modeling

Determination of
evaluation index
Expert brainstorming
Determination of evaluation elements
for each evaluation index
Delphi expert
questionnaire survey
Analysis of feasibility
of questionnaire
Patent evaluation
Structural equation modeling

Expert brainstorming step

Determination and definition of evaluation index

Determination of evaluation elements for questionnaire with satisfactory quantification, objectivity and completeness

Delphi expert questionnaire survey step

In order to collect experts' opinions, it is performed in three rounds with guaranteed anonymity.

In the second and third questionnaire surveys, the previous questionnaire survey is analyzed, so that statistical information on a median and an
interquartile range is provided to the experts as feedback.

Questionnaire survey feasibility analysis step

Content calidity When 30 persons were surveyed, CVR >= 0.33
Verification of reliability A value of 0 to 1, with a greater value representing higher reliability.
Verification of feasibility Degree of convergence and degree of consensus

Steps of construction of structural equation evaluation modeling

Definition of structural equation modeling

Structural Equation Modeling (SEM) is a model in which Confirmatory Factor Analysis (CFA) is combined with path analysis.

A causal relationship between dependent variables as well as a relationship between a plurality of independent variables and a plurality of dependent
variables can be analyzed at the same time.

e1 e2 e3 e4 e5 e6 e30 e31 e32
Evaluation elements
Number of dismissed invalidation trials Number of priorities claimed in a divisional application Number of field responses Number of papers and foreign patents among cited references of a forward citation patent IPC Number of papers/foreign patents among prior documents Change in ownership Presence of earlier publication Number of countries with foreign family patents
Evaluation elements








Right Technology Utilization
Result Evaluation Result

US, EU Patent Evaluation Model

US/EU SMART3 Models applied Multiple Regression Analysis

When analyzing relationships between one dependent and multiple independent variables, multiple regression checks for the following models when
there are K independent variables.

US, EP Patent Evaluation Model numerical expression

(Y : dependent variable, X : independent variable, β : coefficient of independent variable, ε : error)

Multiple regression models use X and Y sets given in the study to produce an optimal β set and ε, which are used to generate the independent variables X,
given in the actual evaluation.

WIPO's Technology Classification

  • Physics/material
  • Chemistry
  • Bio
  • Machine
  • Electric/electronic/IT
WIPO's Technology Classfication

Stanine Ratio

Percentage(%) 4% 7% 12% 17% 20% 17% 12% 7% 4%
Accumulation Rate(%) 4% 11% 23% 40% 60% 77% 89% 96% 100%