Organoid, Yapay Organ ve Biyobaskı Teknolojilerinde Patent Madenciliği ile Teknoloji Tahmini

Abstract

Evaluating technological developments in detail and pursuing these developments is critical process in terms of technology that has been developing since the early days of humanity, making life easier, providing treatment, and reducing time and costs. In today's world where technological developments are being studied with great competition both academically and commercially, choosing the right technological development and transferring investments to the right ideas has become of great importance. That's why businesses and policymakers often use technology forecasting to assess the likelihood of new products succeeding in today's fast-paced and competitive market. Within the scope of this thesis, the patent mining method, which is one of the technology forecasting methods, was applied to the patent data of organoid, artificial organ and bioprinting technologies, which have large markets in health technologies, and to decide which areas should be invested in these technologies was investigated. The international patent classification (IPC) codes are used to select technology areas. Technologies have been evaluated using various technology estimation methods (technology life cycle, technology diffusion rate, patent strength and expansion potential), both individually and in multiple comparatives. In addition, in the thesis study, the ARIMA model parameters (p, d, q values) for each subclass-IPC were tested using values in the range [0, 5] with a piece of code written in the Python (version 3.12.x) program, and therefore each subclass-IPC It was automatically tested using a total of 216 different combinations for the group. This process is designed to determine the best model configuration based on the unique needs of each subclass-IPC. Here, various parameters were compared with the root mean square errors (RMSE) metrics calculated on the test data, and the parameters that automatically gave the lowest RMSE value were assigned as the best parameters. This automatic selection mechanism enabled the selection of the best predictive model suitable for the dynamics of various IPCs, resulting in a significant increase in efficiency and prediction accuracy in the modeling process. This method has played a significant role in optimizing the use of time and resources by increasing automation and efficiency in the modeling process on large data sets. The methods used as a result of this thesis study; It aims to provide a scientific basis for companies and individual investors in any technological field to make investment decision.