以專利探勘探討企業開放式創新策略之研究

To Investigate Open Innovation Strategies of Enterprises based on Patent Mining

楊彩汝(Yang Tsai-Ju)

指導教授 ： 陳灯能

國立屏東科技大學/管理學院/資訊管理系所/碩士(2022年)

https://doi.org/10.6346/NPUST202200021

摘要

由於研發成本的增加以及產品從開發到上市的生命週期逐漸縮短，企業意識到內部研發成本過於昂貴及緩慢，因此開始與外部技術夥伴合作，利用開放式創新策略；開放式創新涉及大量知識的流入和流出，而目前全世界皆是以專利來保護個人或企業的智慧財產，且專利蘊涵了大量資訊，包含重要的技術知識和研究成果；透過專利引用資訊追蹤內部和外部企業之間的知識流動，探索企業與合作夥伴及其他技術領域之間的關係，從知識流動的狀況，就能知道其屬於何種開放式創新。
本研究將使用本研究室所開發之專利成熟度評估模型，獲取企業核心專利，並利用專利引用分析從中萃取專利特徵，以此分析企業開放式創新策略，最後將其結合企業財務指標利用八種機器學習模型來評估企業與開放式創新之間的關係，並利用非線性回歸解釋特徵的重要性。
最後結果顯示，倒傳遞類神經網路的效果在特定的數據集中，企業開放式創新策略的評估效果最好，且利用非線性回歸，可以通過專利引用網路清楚地知道企業知識流動的重要性。

關鍵字

專利探勘 ； 開放式創新 ； 核心專利 ； 企業績效 ； 機器學習 ； 知識流動

並列摘要

Because of increasing R&D costs and shorter product life cycles, enterprises increasingly realise that internal R&D may be prohibitively expensive and too slow. Therefore enterprises start cooperation with external technology partners and utilize open innovation strategies. Open innovation practices involve abundant inbound and outbound flows of knowledge. At present, the whole world uses patents to protect the intellectual property of individuals or enterprises. Patents contain a lot of information, including important technical knowledge and research results. Track the flow of knowledge between internal and external enterprises through patent citation information. And explore the relationship between enterprises and partners about other technical fields.
This study will use the patent maturity prediction model developed by this laboratory to obtain core patents of the enterprises. In order to analyze the enterprise’s open innovation strategy, we use patent citation analysis to extract patent features from patents. Finally, it is combined with corporate financial indicators to evaluate the relationship between the enterprises and open innovation with eight machine learning models.
The results from our model that the effect of the Back Propagation Neural Network is in a specific data set, and the evaluation effect of the open innovation strategy of the enterprise is the best. By using XGBRegressor, we can clearly know the importance of the knowledge flow between inside and outside to the enterprise with patent citation network.

並列關鍵字

Patent Mining ； Open Innovation ； Core patent ； Firm Performance ； Machine learning ； Knowledge Flow

參考文獻

1. 孙涛涛、唐小利、李越（2012 ）。核心专利的识别方法及其实证研究。图书情报工作， 56(04)，80。

2. Abbas, A., Zhang, L., & Khan, S. U. (2014). A literature review on the state-of-the-art in patent analysis. World Patent Information, 37, 3-13.

3. Albert, M. B., Avery, D., Narin, F., & McAllister, P. (1991). Direct validation of citation counts as indicators of industrially important patents. Research Policy, 20(3), 251-259.

4. Azzopardi, L., Vanderbauwhede, W., & Joho, H. (2010, July). Search system requirements of patent analysts. In Proceedings of the 33rd international ACM SIGIR conference on Research and Development in Information Retrieval , 775-776.

5. Bamford, J. D., Gomes-Casseres, B., & Robinson, M. S. (2003). Mastering alliance strategy: A comprehensive guide to design, management, and organization. John Wiley & Sons.