當創新總是發生在不同專業領域的交界之處，團隊成員之間有效的知識互動便成為組織維持其競爭優勢的必要因素，因此為了持續創新，團隊成員必須克服不同專業領域所存在的知識邊界障礙，以便能夠整合團隊成員的知識與專長。在資訊系統的開發過程中，系統分析階段是觀察系統分析人員與使用者知識互動的最佳情境，因此本研究首先探索系統分析情境下影響互動雙方知識分享意願的前因，即創新氣氛與互動雙方的互動投入程度，並檢測這些因素之間的關係及影響。其次，本研究彙整四類邊界物件的內容，即語法的、語意的、實務的與隱喻的邊界物件，探討不同類別的邊界物件之間的關係及對專案創意的影響，並更進一步探究互動雙方的知識分享意願對各類邊界物件的影響。 研究樣本來自258位學生所扮演的系統分析人員及258位業界使用者，研究結果發現創新氣氛會同時影響系統分析人員與使用者的知識分享意願。當系統分析人員的互動投入程度越高，使用者的互動投入程度也會提高。使用者的互動投入程度會正向影響系統分析人員及使用者自身的知識分享意願；但是系統分析人員的互動投入程度卻對使用者的知識分享意願產生直接且負向的影響。 此外，系統分析人員的知識分享意願只對隱喻的邊界物件有正向影響，而使用者的知識分享意願卻對這四類邊界物件皆無顯著影響。在這四類邊界物件之中，只有實務的邊界物件對專案創意具有正向的影響，而隱喻的邊界物件則對於語法的、語意的、實務的邊界物件三者具有正向的影響，因此隱喻的邊界物件可以藉由實務的邊界物件對專案創意產生正向的間接效果。出乎意料之外，語意的邊界物件對專案創意具有負向的影響。在實務方面，本研究建議組織應該藉由增加實務的邊界物件和隱喻的邊界物件，以提高專案創意嶄露的可能性。 / As innovation happens from the boundaries between different specialty domains, effective knowledge interaction among participants has become a competitive necessity for organizations. In order to improve innovation, it is necessary for participants to deal explicitly with the interpretive barriers and further synthesize their expertise. During the development of information system, most knowledge interactions, which are conducted by systems analysts and users, can be observed in the phase of systems analysis. Thus, this study intends to explore not only factors that may affect participants’ intention to share knowledge, but also types of boundary objects in the context of systems analysis. As an innovative organizational climate and participants’ interaction involvement are identified as antecedents of participants’ intention to share knowledge, the relationships among these factors are studied. Besides, while four types of boundary objects are summarized, their effects on project innovativeness are discussed. Then, the influence of participants’ intention to share knowledge on the occurrence of four types of boundary objects was further examined. The results, based on data collected from 258 student analysts and 258 users, indicate that an innovative climate positively and directly influences both systems analysts and users’ intention to share knowledge. While systems analysts’ interaction involvement has a strong influence on users’ interaction involvement, which is measured by responsiveness, perceptiveness and attentiveness, users’ interaction involvement also has positive effects on both systems analysts and users’ intention to share knowledge; unexpectedly, systems analysts’ interaction involvement is negatively related to users’ intention to share knowledge. In addition, while systems analysts’ intention to share knowledge only influences the occurrence of metaphoric boundary objects, users’ intention to share knowledge does not demonstrate any significant effect on four types of boundary objects. Out of syntactic, semantic, pragmatic and metaphoric boundary objects, only pragmatic boundary objects had a strong positive effect on project innovativeness. As metaphoric boundary objects are shown to have positive effects on all other three types of boundary objects, metaphoric boundary object affects project innovativeness indirectly through pragmatic boundary objects. Surprisingly, semantic boundary objects negatively affects project innovativeness, although not very strongly. A practical implication is the possibility of increasing project innovativeness by enhancing both pragmatic boundary objects and metaphoric boundary objects.