Expert Systems

Expert systems are artificial intelligence applications designed to emulate human expert decision-making in specific domains (Á. Rocha & S. Anwar, 2019; T. Todorov & J. Stoinov, 2019). These systems solve complex problems by reasoning through knowledge bases, often represented as if-then rules (Á. Rocha & S. Anwar, 2019). They are applied in various fields, including milk quality monitoring and animal health (T. Todorov & J. Stoinov, 2019), as well as information management, organizational models, and software systems (Á. Rocha & Lima, 2018). As technology advances, expert systems are evolving to demonstrate superb decision-making skills and conform to social norms for expertise, behaving more like human experts (Papageorgiou, 2021). This development has implications for various fields, including bioinformatics, and raises questions about the concept of "expert generalist" (Papageorgiou, 2021). The continued research and development of expert systems contribute to their growing capabilities and applications across diverse domains.

Reference:

1. Rocha, Á., & Anwar, S. (2019). The journal of knowledge engineering special issue on WorldCist'17—fifth world conference on information systems and technologies. Expert Systems, 36. https://doi.org/10.1111/exsy.12414
2. Todorov, T., & Stoinov, J. (2019). Expert System for Milk and Animal Monitoring. International Journal of Advanced Computer Science and Applications.https://doi.org/10.14569/ijacsa.2019.0100604
3. Rocha, Á., & Lima, S. (2018). Expert systems: The journal of knowledge engineering special issue on WorldCist'16 ‐ 4th world conference on information systems and technologies. Expert Systems, 35. https://doi.org/10.1111/exsy.12260
4. Papageorgiou, K.G. (2021). Expert Characteristics: Implications for Expert Systems. Advances in experimental medicine and biology, 1338, 155-164 . https://doi.org/10.1007/978-3-030-78775-2_18

Backward Chaining

Backward Chaining adalah salah satu metode pengontrolan inferensi berbasis aturan. Pelacakan dimulai dari tujuan. Selanjutnya dicari aturan yang memiliki tujuan tersebut untuk kesimpulannya. Selanjutnya proses pelacakan menggunakan premis untuk aturan tersebut sebagai tujuan baru sebagai kesimpulannya.

C0ntoh:

Aturan:

R1.      Jika A dan B, maka C

R2.      Jika C, maka D

R3.      Jika E atau F, maka G

R4.      Jika D dan G, maka  H

R5.      Jika P atau Q, maka S

R6.      Jika S dan C, maka R

R7.      Jika R atau H, maka X

BUKTIKAN KEBENARAN “X”, JIKA;  A,B,E,F,Q =  TRUE dan  P = FALSE !

Penyelesaian:

NO	ATURAN	ATURAN BARU
R7	JIKA R ( - ) DAN H ( - ), MAKA X ( - )	-
R6	JIKA S  ( - ) DAN C ( - ), MAKA R ( - )	-
R5	JIKA P ( F ) ATAU Q ( T ), MAKA S ( T )	R5,  S = TRUE
R1	JIKA A ( T ) DAN B ( T ), MAKA C ( T)	R1, C = TRUE
R6	JIKA S ( T ) DAN C ( T ), MAKA R ( T )	R6, R = TRUE
R4	JIKA D ( - ) DAN G ( - ), MAKA  H  ( - )	-
R2	JIKA C ( T ), MAKA D ( T )	R2, D = TRUE
R3	JIKA E ( T ) DAN F ( T ), MAKA G ( T )	R3, G = TRUE
R7	JIKA R ( T ) DAN H ( T ), MAKA X ( T )	R7, X = TRUE