基於科學與技術知識缺口的技術機會分析：以廢棄鋰電池回收技術為例

企業的動態能力始於技術機會的發掘與掌握。在技術機會分析的研究路線中，有一條考量科學與技術的知識缺口阻礙了技術的發展，藉由找出知識缺口而發掘技術機會。在此研究路線中，雖有過去文獻基於新穎的深度學習分群法確認出科學與技術的知識領域，但後續仍需高度仰賴專家解讀各群的知識。為減輕專家的判斷負荷，本研究提出可藉重主題建模的方法萃取文件集的知識主題。本研究提出一個結合科學與技術知識而發掘技術機會的架構，架構中主要結合了基於深度學習的分群法、主題建模、與餘弦相似度測量。有鑑於發展鋰電池回收技術是我國推動2050淨零排放路徑的關鍵計畫，本研究將所提架構應用至此技術。研究結果顯示本研究的架構確實能有效地找出技術機會。實作後發現回收與修復陽極材料以及鋰電池的電池管理與監測兩領域的科學知識較少應用至專利技術，是潛在的技術機會。

An enterprise’s dynamic capabilities originate from analyzing and seizing technology opportunities. In the study of technology opportunity analysis, one research path discovers the technology opportunities by identifying the knowledge gaps between science and technology, considering that such gaps may impede technological development. Previous studies on the path have introduced a deep learning-based clustering approach to identify knowledge fields in science and technology. However, one limitation exists in that the interpretations of each field heavily rely on experts. To alleviate the burden on experts, this study finds that a topic modeling approach could be used to extract knowledge topics from corpora. Therefore, this study proposes a framework integrating science and technology knowledge to uncover technology opportunities. The framework combines approaches of deep learning-based clustering, topic modeling, and cosine similarity measures. In view of the fact that developing lithium battery recycling technology is a key project for our country to realize the goal of 2050 net-zero emission, this study selects the technology as an empirical case to implement the proposed framework. The results demonstrate that the proposed framework can effectively identify technology opportunities. For the empirical case, the findings indicate that knowledge in two scientific fields, i.e., recycling and recovering anode materials and the managing and monitoring of spent lithium batteries, needs to be applied more to patented technology, implying potential technology opportunities.