隨著經濟的轉型與科技的進步，長久以來企業的競爭利基已經由有形資產轉變為智慧資本，而過去智慧資本的評估指標中大多採用數值性之資訊或問卷的方式獲取，嚴重忽略了文字資訊對評估的重要性，有鑑於此，本計畫將提出以社群媒體探勘技術擴展智慧資本之指標，透過技術間之相互整合，可以獲取經理人的動態管理能力、員工的社群網絡關係、董監事的政治連結關係、假評論的偵測、顧客的關係網絡、企業未來趨勢的掌握能力、符合國際標準的指標等，並將獲取的結果匯入到人工智慧技術，進一步探究各個指標對企業營運表現的影響，最後再利用數學規劃模型提供企業應該改善的方針與構面，進而達到永續發展的目標。
With the transformation of economy and improvement of information technology, corporate’s competitive advantages had been switched from physical assets to intellectual capitals. In the past, the performance indicators of intellectual capitals were collected mainly from financial reports (i.e., numerical information) or questionnaires, and ignored the textual information. However, it is widely recognized that the textual information can be a complement to numerical information. Based on this viewpoint, this study tried to extend the indicators of intellectual capitals by utilizing social media mining and text mining technique. The manager’s dynamic managerial capability, employees’ business relations, board of directors’ political connection relations, fake reviews detection, consumers’ sentiments, the ability to reach the top level in information technology were obtained by joint utilization of fusion methods and examined the impacts of each indicator on performance evaluation. The analyzed results were then fed into artificial intelligence (AI)-based technique to construct the forecasting model. To extend the applicability of the AI model, this study further extracted the inherent decision logics from AI model. The decision logics can be examined by users will improve the acceptance of the AI model. The proposed model not only poses superior forecasting quality, it also provided guidelines for managers to allocate the scare resource to suitable places as well as reached the goal of sustainable development.
