資訊管理學報
盧芸玲;劉瑞瓏;
頁: 33-54
日期: 2004/07
摘要: 隨著知識經濟時代的來臨，知識之有效累積與應用已成為企業永續經營的重要關鍵。然而現今知識管理與資訊科技尚存在著差距，其中又以企業動態知識之自動分類為主，此乃因企業之知識往往相當龐雜，且具變動性與累積性。本文從知識管理與資訊科技兩者之整合角度出發，探討並縮短此差距，讓資訊科技能在企業內輔助知識管理活動之推行。我們提出一個合作式多代理人知識分類與管理模型，將企業員工與知識緊密結合。多代理人透過協商來對輸入之知識文件與需求進行妥善分類，並藉以確定負責此相關知識管理活動之代理人。各代理人即可各司其職，互助合作，達到知識之誘導、儲存與共享，有效支援整個知識管理之活動流程。我們並實際驗證此模型之知識分類技術。實驗證實：在階層式組織分類架構下，該技術能成功地自動擷取知識之適用場景，各代理人即可據此進行合作，提高知識自動分類之準確度。
關鍵字: 知識分類;合作式代理人;知識管理;
Abstract: Knowledge management (KM) has become a critical activity in business administration. However, there exists a significant gap between knowledge management and information technology: automatic classification of dynamic business knowledge, since business knowledge is often dynamic and accumulative. This paper aims to explore and reduce the gap so that information technology may really assist businesses in managing knowledge. An agent-based collaborative knowledge classification and management model CKCMA is proposed. The agents collaborate with each other in order to closely link people to knowledge, and vice versa. They negotiate with each other to classify input knowledge documents and requests, and accordingly identify which agent should be in charge of the related KM activities, including knowledge elicitation, accumulation, distribution, and sharing. An experiment is designed to empirically evaluate CKCMA's multiagent knowledge classification technique. In the experiment, the feasibility and contributions of context-based automatic knowledge classification are verified.
Keywords: Knowledge Classification;Collaborative Agents;Knowledge Management;
瀏覽次數: 10368 下載次數: 79
引用 導入Endnote
頁: 33-54
日期: 2004/07
摘要: 隨著知識經濟時代的來臨，知識之有效累積與應用已成為企業永續經營的重要關鍵。然而現今知識管理與資訊科技尚存在著差距，其中又以企業動態知識之自動分類為主，此乃因企業之知識往往相當龐雜，且具變動性與累積性。本文從知識管理與資訊科技兩者之整合角度出發，探討並縮短此差距，讓資訊科技能在企業內輔助知識管理活動之推行。我們提出一個合作式多代理人知識分類與管理模型，將企業員工與知識緊密結合。多代理人透過協商來對輸入之知識文件與需求進行妥善分類，並藉以確定負責此相關知識管理活動之代理人。各代理人即可各司其職，互助合作，達到知識之誘導、儲存與共享，有效支援整個知識管理之活動流程。我們並實際驗證此模型之知識分類技術。實驗證實：在階層式組織分類架構下，該技術能成功地自動擷取知識之適用場景，各代理人即可據此進行合作，提高知識自動分類之準確度。
關鍵字: 知識分類;合作式代理人;知識管理;
Collaborative Agents for Knowledge Classification and Management
Abstract: Knowledge management (KM) has become a critical activity in business administration. However, there exists a significant gap between knowledge management and information technology: automatic classification of dynamic business knowledge, since business knowledge is often dynamic and accumulative. This paper aims to explore and reduce the gap so that information technology may really assist businesses in managing knowledge. An agent-based collaborative knowledge classification and management model CKCMA is proposed. The agents collaborate with each other in order to closely link people to knowledge, and vice versa. They negotiate with each other to classify input knowledge documents and requests, and accordingly identify which agent should be in charge of the related KM activities, including knowledge elicitation, accumulation, distribution, and sharing. An experiment is designed to empirically evaluate CKCMA's multiagent knowledge classification technique. In the experiment, the feasibility and contributions of context-based automatic knowledge classification are verified.
Keywords: Knowledge Classification;Collaborative Agents;Knowledge Management;
瀏覽次數: 10368 下載次數: 79
引用 導入Endnote