우리나라 주요산업의 기술특허정보예측모형의 구축

Abstract

The purpose of this study is to set up optimal model to forecast technology patent information in four major industries in Korea. The major procedures are as follows: first, construct the model which extracts time-lag between input and output indicators on R&D activity; secondly, examine various models to forecast the values on technological information of patent by R&D; thirdly, extract the optimal forecasting model in each industry by testing the precision of values forecasted. The major findings are summarized as follows: Time-lag between input-output indicators showed quite different result in patent and utility applied, but in the case of patent granted there are time-lag of about 5 years. In general, the extrapolation model is more precise than the causal model in forecasting R&D outputs.