發展以知識本體為基礎模擬專案團隊之組成

Developing an ontological knowledge model for composing project team members

許志瑋(Chih-Wei Hsu)

指導教授 ： 戚玉樑

中原大學/商學院/資訊管理研究所/碩士(2010年)

https://doi.org/10.6840/cycu201000903

摘要

企業為了快速因應外界環境的變動，以達到企業本身的最高收益及永續經營，成立專案團隊是最常見的方式，本研究提出以知識本體(Ontology)為核心的專家系統(Expert system)，利用過往與專案團隊績效相關之文獻回顧，從中找出影響專案團隊人員績效的相關因素並與知識專家合作以知識工程分析的方法，透過問題領域界定分析其領域特徵，藉由領域特徵的概念化進行知識擷取、知識塑模及知識表達等程序，建立開發專案團隊人員組成之知識模型；本研究將利用 Protégé 做為本體知識庫之開發工具，建立其相關之概念、屬性以及描述邏輯的知識庫系統，並使用Jasa Server Page 、Jena API及 Protégé OWL API 等技術開發網頁介面，依據本體架構與規則限制進行推論找出專案候選成員，在選出專案候選人後，專案經理可利用不同的排列方式列出整個專案團隊成員的推薦名單，目前的排列方式有二：(1)依成本排列(2)依績效排列(認知風格、人格特質、年齡接近者)。在依據模型定義的架構下給予實例並進行驗證後，其結果顯示，本研究確實能成為專案經理人在挑選成員時之重要決策依據。

關鍵字

語意規則 ； 知識本體 ； 專案團隊 ； 團隊績效

並列摘要

In order to respond to rapid changes in the outside environment, to reach the goal of the highest revenue and sustainable management, the most common solution enterprises would choose is forming a project team. This study proposed to use the expert system with the ontology being the core, through reviews of previous literature related to project teams’ performance, to explore the factors which influence project team members’ performances, and to cooperate with knowledge experts to apply knowledge engineering method, to analyze domain features by problem domain definition, and to perform procedures such as knowledge retrieval, knowledge model building, and knowledge expression by conceptualization of domain features, so that the knowledge model of composition of development project team members. This study used Protégé as the development tool for ontology knowledge base, to build an ontology system of the related concepts, properties, and descriptive logic. And technologies such as Jasa Server Page, Jena API, and Protégé OWL API were used to develop the interface for the webpage. Inferences of project member candidates were made according to the ontology structure and constraints. After the candidates were picked, a project manager could obtain a recommendation list of project team members sorted sing different methods. Currently there are two sorting methods: (1) sorting by cost and (2) sorting by performance (people with similar styles, personalities, and ages). Real cases were used for verification under the structure defined by the model. The result showed that this study indeed offered project managers some key references to pick their team members.

並列關鍵字

Team Performance ； Ontology ； Semantic Rule ； Project Team

參考文獻

吳慧卿(民91)，團隊衝突對選手滿意度之影響初探，中華體育季刊，16(4)，頁68-74。

李弘暉(民 96)，高績效團隊管理的理論基礎─團隊理論模型綜述，中國行政評論，6(4)，87-104。

林詩穎和張秀華& 周惠文.(民96)，團隊多元性及團隊衝突對團隊績效的影響，電子商務學報，9(2)，頁353-376。

戚玉樑(民95)，協同知識擷取與知識表達程序於建構本體的概念架構，資訊管理學報，13(2)，頁193-215。

張翊翔(民93)，團隊成員人格特質組合對團隊效能影響之研究－以團隊互動過程為中介變數，中原大學企業管理研究所碩士論文。