“…In the past decades, many KA systems, e.g., ETS (Boose, 1984(Boose, , 1985, NeoETS (Boose & Bradshaw, 1986), AQUINS (Boose & Bradshaw, 1987), KITTEN (Shaw & Gaines, 1987), RuleCons (Davis, 1987), MOLE (Eshelman, Ehret, McDermott, & Tan, 1987), KSSO (Gaines, 1987), KRITON (Diederich, Ruhmann, & May, 1987), EMCUD (Hwang & Tseng, 1990;Hwang & Tseng, 1991), KADS (Wielinga et al, 1992), MCRDR (Kang, 1996), KAMET (Cairo, 1998), MedFrame/CADIAG-IV (Boegl, 1997;Kolousek, 1997;Leitich et al, 2001), (Pan, Zheng, Zeng, & Hu, 2002) have been developed to rapidly build prototypes and improve the quality of the elicited static knowledge of well-known objects by domain experts. However, most of them cannot be used to construct the dynamic knowledge due to the limitation of the static attribute set of the static grid in a dynamic environment.…”