The aim of this paper is to present the concepts about Type 2 Approximate Functional Dependency - AFD 2 (relating to the correlation between atributes of the relational file) and the outliers with functional dependency. The paper also gives some properties of ADF 2 and algorithms for finding the outliers with functional dependency. | ’ Tap ch´ Tin hoc v` Diˆu khiˆ n hoc, , (2007), 80—85 ı e e . . a ` . ’. ´ ’ ` ˆ ` ˆ ˆ PHU THUOC HAM XAP XI VA PH` N TU NGOAI LAI A . . . . ´ ˆ ` ˆ ´ DOI VO I PHU THUOC HAM . . ’ ˜ ´. VU DU C THI1 , PHAM HA THUY2 . . 1 Viˆn e . Cˆng nghˆ thˆng tin, Viˆn Khoa hoc v` Cˆng nghˆ Viˆt Nam o e o e e e . . . a o . . ’m to´n Nh` nu.´.c o tˆm Khoa hoc v` BDCB - Kiˆ a e a a . a 2 Trung Abstract. The aim of this paper is to present the concepts about Type 2 Approximate Functional Dependency - AFD 2 (relating to the correlation between atributes of the relational file) and the outliers with functional dependency. The paper also gives some properties of ADF 2 and algorithms for finding the outliers with functional dependency. ´ ´ ´ T´m t˘t. B`i b´o gi´.i thiˆu vˆ phu thuˆc h`m xˆ p xı loai 2 (loai phu thuˆc h`m xˆ p xı liˆn quan o a a a o e ` e o a a ’ . o a a ’ e . . . . . . ` ´n su. quan gi˜.a c´c thuˆc t´ cua mˆt quan hˆ) v` phˆn tu. ngoai lai dˆi v´.i phu thuˆc ´ o ’ ’ u a o ınh o e a a o dˆ e o . . . . . . . ´ ´ ´ h`m. B`i b´o c˜ng tr` b`y mˆt sˆ t´ chˆ t cua phu thuˆc h`m xˆ p xı loai 2 v` thuˆt to´n x´c a a a u ınh a o o ınh a ’ o a a ’ . a a a a . . . . ` ´ o a dinh phˆn tu. ngoai lai dˆi v´.i phu thuˆc h`m. a ’ o o . . . . ´. ˆ GIO I THIEU . ´ a Kh´i niˆm phu thuˆc h`m xˆ p xı (Approximate functional dependency) v` ph´p a e o a a ’ a . . . ´ ´p xı d˜ nhiˆu t´c gia dˆ cˆp dˆn v` dung ` a ’ a ’ ` a e a e . ph´t hiˆn c´c phu thuˆc h`m xˆ a e a o a a e . . . . . ´ . ’ ˜ ` ’ u o a e e trong nhiˆu b`i to´n phˆn l´.p cua data mining ([1, 2]). Nh˜.ng phu thuˆc nhu. vˆy biˆ u diˆn e a a a o . . . . phu thuˆc tu. nhiˆn gi˜.a c´c thuˆc t´ trong mˆt quan hˆ c´ ch´.a mˆt v`i h`ng o . e u a o ınh o e o u o a a su . . . . . . . ´ ’ o c´ sai s´t ho˘c khˆng thoa luˆt (goi l` phu thuˆc h`m xˆ p xı loai 1). Mˆt tru.`.ng phu o o a o a a o a a ’ . o . . . . . . . . ´ o o o ınh a u u u o o . thuˆc xˆ p xı kh´c l` c´ nh˜.ng nh´m .