功能相依性(Functional dependency)和多重值相依性(Fuzzy multivalued dependencies)在資料庫中扮演很重要的角色,它們的功能在於消除資料庫中重覆的資料以及維持資料分割或合併的一致性。在本篇論文中,將結合幾位學者的論點提出新形式的模糊多重值相依性,並證明其推導規則(Inference rules)的正確性(Soundness)。 1997年S.jyothi 和 M.syam Babu 提出模糊多重值相依性(Fuzzy multivalued dependencies)的定義。同時,1999年,S. Ben Yahia、H. Ounalli 和 A. Jaoua 提出動態模糊功能相依性(Dynamic fuzzy functional dependency, DFFD)的概念,在其論文中運用模糊邏輯(Fuzzy logic)的概念以Yager所提出的S-norms(模糊聯集的一種表示方法)之Lukasiewicz模糊關係計算法則,提出動態模糊功能相依性(Dynamic fuzzy functional dependency, DFFD)的概念。 在本篇論文中,將以模糊邏輯(Fuzzy logic)的概念及S.jyothi 和 M.syam Babu所提出的模糊多重值相依性(Fuzzy multivalued dependencies, FMVD)這兩個概念,提出動態模糊多重值相依性(Dynamic fuzzy multivalued dependencies, DFMVD),並推導其推導規則(Inference rules)的正確性(Soundness)。
ABSTRACT Functional dependency (FD) and multivalued dependencies (MVD) play an important role in the classical relational database. Since it is not only a way to eliminate the redundancy in the database but also make data consistent. In this thesis, we will combine several scholars’ ideas to present a new form of multivalued dependencies and try to prove they are consistent with inference rule of classical database. S. Ben Yahia, H.Ounalli and A.Jaoua proposed the idea of dynamic fuzzy functional dependency (DFFD) using the fuzzy logic of Lukasiewicz and Yager’s S-norms as the basis in 1998. Likewise, we expand the idea of dynamic fuzzy functional dependency by combining the fuzzy multivalued dependencies proposed by S.jyothi and M.syam Babu in 1996 with fuzzy logic to present the new form of multivalued dependencies called “dynamic fuzzy multivalued dependencies (DFMVD)” and prove the related inference rules are soundness.