!"#$%
&'()(*+,-
!" #
$%& #
! "
2
0
1
1
2
Bài tập lớn môn Xử lý ngôn ngữ tự nhiên
Mục lục
H0ST.BFA4U+&T+O*+8V6+=+30@WU*F
X5&VYZ+6+H00+6+4[0F+&B,WJF\A2]
3+F+620ST.BFA5^_O8++&=H0ST.BFA4U0P
B<N0+3,60!2+&`+&F0SJR6A+6+5aO6'
RST.BFA2A.b5R 4"WC++aB_OH00c./06
A+6+4U+d0Sd<3,A0F+&B.BFA+&T+F+&B+6++6
200060/2]F\,WJF\0ST.BFA5H0F+&B,WJF\
+6+0POe+TVf8O`5g<C+0cO*++h0c0H,N<C+ViO*+O6'
2O60%4"0PB0,"F+&B,V+0+6+5
AO+6+T41<;O0]+6+,"*+I0@+h0c0,+6+,"
O6+&X,+h0c0I+4*5
00O6'.BFA8++&=TOc04*0H00+6+F+&B
.N.I.BFA88j4.IHX+7&0k&C+"+&+c0
0k+"O9+&.BFAO00Oc081+-0+&N04PV68+A&5\.
0@,C4"F0081+-0+&N04P0YONNO/04-00/+=0@0
X5R0O/04-0F0l4,0081+-0+.0X
4&+&0]0/+=500+6+8]O6+&X+4h+'
0XV67&4=4"0Y0081+-0,4&0081+-0ON5
R0+&+c04P0P+=FNV.H00.d<++&XH0%0P+=F
Ol2AB00+&X*..BFA555O0XV6'.
&4[0V+3O6'00A+l5$'+3HZ0c,"+
A+&+c0+&0ST.BFADmF>.>GS0,>&G+<S>E&4X,N<
+8.BFADG+)EFO+&+1OZ0c5R0++TZ
0c,00+7++,"&-+A1+I,A0Z4U4[08./,+
4[0BV3+2]&C+2+&`(0123+4HOI?&555
)*+Sl<+2+&`+&8.BFA,_<]4[0Wn+43
F 00 < + o>W+ RFSSp0+qH o>W+ ^O&r+qH , o>W+
R+>&r+q5
&Z+3N4U0P&C+"+06+&4"+81FN8,_<]00
Z0c0@Us )H+&008k+-AOT)sT00,A
00+6+V+3H00<Z0cV`05G!8./RST.BFA,
+&I+4*0@+h0c0F&C+8h<3,4>OFI"F[-0VF+&B,
WJF\H+V6+=2Z4[0&x0k&C+".I+6+V04[0
F+&B.N.I,_<]57O0-0]+&00+6+4[0F+&00
0ST.BFA+'8FN+&Sl0j0%4[0+h0c0.N.I,_<]5
AH00+h0c04U8./06A+6+,2]F\A+l06
,_C+XH,-./00A+lSJ./+S.>5Z4P0Y+d0SdF
002]F\F;.BFA,_<]H00C800060/V0cH0k.BFA,y
+d0SdxO.N.I,_<]5Rj+00P00]+7+81FIH+'OV3O
+FAH00060/+&-0F`0+6+xOO/04-0+lVZH8+A+&+c0H
&23+4+&d0+38+&Z00;.BFAV=5
$N+d0+34PH,C4"4e+&FFO+30j+0P+=V+04[0
B+6+B-0+a00;+FA,_<]P05R0;.BFA
8]4[0WJF\+34=X.!0P+=0PB060/+d4*
+&[j8+&,A08+A+&+c0,V+0+6+5Mz&H0j+
8]=&z<]0C+0@.BFA,_<]H=&z004e0+&0@00.BFAFI
4=0P+=0P4[0B888F70+3+5
$A0V+0+6++a00;.BFA,_<]+&00+h0c0$A+
O0f00f8].d,BV3+2]Z0c,",_<]P0H,"
.BFA,_<],00Vt+7+WJF\4U4[08++&=+&Z+3N5ZH
B,_<]+3$A+FI0PB4e0+&&Z0@P50P+=7+C
4[0SdV0<A+,"Oe+V-88H0j88,B88+3$A++&00
,_<]S,N006B8h<3+&Z+3N+3KH+385$7
+'B4e0+&]T+34300Vt+7+V8.BFA,_
<]H+08]0PBV\+7+ON4=0P+=+7./4[0B+3
0@+3$A+0%]23+4[0B8c0+I8+&+3$A+5
93:;<=>-*?@A-
932*B+(+,CDE-+(+()
!"#$%&'(
$A0SJ./0ST.BFA,I+4*0@O*++h0c04U4[08++&=
0P+=+'O4[0+a<C+V'O*+R^G0YBOy8![8,NO/0
4-0+'OV3OON4[0`F+&+c05Sv0PBOSl4=400+Z
0-OyCL+VPU=K+V7B()7WKPEDCEX-3
:*ON0@Oy'8/+*0,V8IO,203H0P+=4l,N
A+le04l,NX.!5$-./,N.BFA0@O*+06+H2+&'
8.BFA+'O&4[0O*+F7+[+c0+4[0]O,O!+T
,!8- f0H4l,NA+l+'&C+ONH+&N0V0"0P<C+0cO*+
0<*F78V3I00%7&4[04"200<0+0-5
-B./0@Oy0P+=44[02SdFZ243O/04-0+'O
V3O5$NO*+0<*8/+&0<]+&'O+-T06++'F7++&ZV60P
+&HOe0.!FON4l,N+5
RP+=2064IV8+&+c00P&C+"Oy4[0FC&
V68]Oy0%0P+&H0P+=FONHB-0FI+O+XH
4e0<A+FV8./00Vt+7+.d+&Z+lVZ5G4PF68]0P00+Z
0-,00O400Oy4W>OWn+HV6+O+X5
POFIH8.BFA+d0&0P+=0FO*+2+&'W04Oy+a
00G+m&>S>HSJ./00Vt+7+S}0P`0OH7.IH+lVZH
81I555,00Vt+7+4[08++&=<TZ0c+&-+A1+I
)I?&1+ID>+&F>+m&VEH00+7++.+&"D>>&0
F&+OEH2I8F7+&F>&>.0+E555
0P+=Wn+43O*+Sl<+0-4l,NZ0c,"8.B
FA
? )*D0FSSp0+E('OO*+WID81FIE+aO*+Oy
.BFA,O*++&00FN80+&N05
? )+,-D&>&>SSE('OO*+WI;2+aO*+Oy.BFA
,O*+<3.d40P+&+d0
POQR) >5
m><S+>+OI55500Z0c,"0ST.BFA&4X,B_O
HX++T&x0P+=FO`FI+6+.N.I.BFA0P0C+&j05
+&Z+d0+3S_O8++&=HX+,y.!00A+lF
+&BT.I,_<],+7O+&-0k0PWN.!+XWZ5a4P
X+0P+=+&x00S]89O8.BFA,_<]0P+=0P+&
+OI0&C+"FS,N00S]89O8.BFA+&"
+lV05Z+0%0P+=+C&x00Vt+7+8.BFA
,_<]8c0+I8"S,N00Vt+7+8.BFA+&"+l<T,'
8]+d0A+&Z.BFA,_<],l4UT.I80C+&j0,0P+-OX
DprrE5
)*+,-./0<+V8.BFA,_<]HV81+-000<<
Z0cV`0H+0P00+6+S(
? oS+&>SSFO*+<AFZ24344q
? oS+&>SSWC+A0P+=.+3)Z+&Oq
? oRW0P+=_0]O*+Sl0c44q
? o)ZFO*+Z+l4"0W+dZ+&Oq
^V81+-000+6+2+&`HA+l08]4&00
SF10/+=O+-00OI(
? o3/+)Z0P+=1&O*+Sl<A44q
Mz&T410PSd81+-0SF7TOc04*05:=4I+4[0V_
,708]0PB06+&'Z0c,"+&-+A1+I+Z+35
POQR) >6
81+-0+lVZ,81+-000FZV3+5R0<N00kFISvV80],_
<]44@+a+7800,_<]H,-./81FN8,_<]5
)/0+Z0l0!0@>W+)+XF4XFlA2]H+AH,
4e0+&4=+&'.,+'OV3O00+78[8&*FN0@00,_<]5G4PH
00Vt+7+0-0@>W+)0P+=4[08181&+00AO,/
O0j+d0AVWJF\V8,_<](FI+6+O0j0P+=+&-0
&,FI81+-04[0+d0A<T0j5
R0FI+6+4[0+&-0&0P+=F(
? BP(*Y(Z•]SJH4[0FZV3+,NO{,_<]F+7800U00
++0V8+&+c04[0+d0A+&Z00U0@O{,_
<]5P0H0P+=]SJ&x00U+c,N00+a
VHO{O*++aV0P2A,NO*+0@4"0/+=4P5
? BP-TZg41]SJ&xO*+,_<]4[0U,N+a+a
WC+A+&,_<]4P5
? BP-*?[-()NZg41,NO{,_<]+'O+C000{+aH0{
+a4P+*0,"O*+Fu,d04P,.4P,A0+'OV8,_
<]4[0+d0A+&Z00VAO4[0U0O{,_
POQR) >7
2
0
1
8,9:29;<C+0cO*+O6+=+V05R0A
+l81FI,_<]+XFO,A0,NO*+ <=2 DF>&
F&+OE57++4P4[000C8O*++78Oy4=8/0,/0,A0.I`05
78Oy<;OO*++7800+d0+=0PU4[081FN8+&N00P
.IDWHET4PWF+d0+=4[081FN8HFUD81FN8E4[00
P5$N00C0ST,7HVO*++d0+=4[000C80A+lHPSv
0lfS&O*+O+`0+a+784+IOy,WI+d0+=ON4P,
O*+81FN851FN8,_<]F<+,40PB<N08++&=
3+Sc02+&`OZ10@3.Sd8++&=OIOv410@
00+6+Z+&d0+35
1! 2 ! .#/0)*
345678 .9 :7
)*+888+lVZ81FN8,_<]4[04"V=<TO*++a4=
0@4"0P810C84[04"WC+588SJ./O*++a4=,NO*+
0C+&j04]5a4=0P+=.I4[0..+&ZO*++78[8+FA
POQR) >8
2
0
1
1
2
<C+Vi2ABu+Z09V=0H8H€
R10@4"4[0+h0c0+O*+810C8HP0F+h0c0+
O*+OIDV4PO*+Sl0@4"0P+=+*0O*+,j+0@01810C8E5
•]+7++'OV3O0@4"+&Z+a4=0%;O0P8(+'OV3O0@
4"4D0@4"FE,Sd+&"F+&`Sl0@0@4"+&Z015d0+3HP+&]FXH
001•S(+NOc04*+'+FASv8!,N0@4"4U0†)*+01
•,74[0+&]FX0O{0@4"&Z<A+5&+&X[84]C+H
+&`Sl0@O*+0@4"FSlD+SC+E00+a+cH+&.S0+aH4[0
+'O+C+&+FA„…5
+c0@]+7+0P+&0AOF+&"00+SC++'O+C
+&Z01„…5$N8]+7+H0j+0P+=0Y&&xO*++FA4"078
430@,"B4*,7+0P,jHB4*,7++1O"OHB4*,7+8
S+Tj+FH8![8,N0@4","B4*,7+H00S,7+Sl,+dZ
V6Tj+F5
POQR) >9
2
0
1
1
2
Bài tập lớn môn Xử lý ngôn ngữ tự nhiên
I
I
0
E
J
„…5g418n8+-+h4[0+d0A+C+0]00+FA0P
S}GH
I
E
F8n84Sd8![80@+FA,N0@4"ˆH,
0
FSlFWC+A
0@+a0@4"V+&+FA5
6OH0j+V6++I<C+0c]+7+4+074=
+'O&8n844*8![8
I
E
0@00+FA000Fu,d0O*+004*0F785
,4PHO*+8n84,74[04<x+<T0ZH,S
4PA+l4[0CFA+&Z[800+FA5R00Z0P+=8]
+XWZB+&`Sl+-0[8<x+000+FA5
R]00+3807Z0&x4[0FO<_+5:=+&4"4PH
,NO*+8n8+4jH,NB4"+4@‰8H0P+=]+3+&xB
,_<]+&Z,"0@4"V6<XWC+A+&B,_<]
I
I
KEJ
ED
HT4P
∑
∈
=
#
I
#EK ‹
F+&+&<'0@
I
E
20ST.B
POQR) >10
2
%
043
V,N+NO*+j+F54PHU0@j+4[00%
I
5:Sl008N
8881FI,7SJ./<=.,_<]T.I81H,,700
010%4[0<=..N.I815)*+,-./,"0123+44[0
OI+&'5
)*+888V]+.!4=CFAO*+0123+481FI2
xOT03F[028EL„…5R3F[0SvV=O+&W>OFA+C+0]00
POQR) >11
lúa mì lúa mì
nông trại nông trại giạ giạ
thương nghiệp
thương nghiệp xuất khẩu
xuất khẩu
nông nghiệp
tấn tấn
nông nghiệp
lúa mìlúa mì
lúa mì
lúa mì
lúa mì
lúa mì lúa mì
lúa mì
Hình 1: Một ví dụ về cây quyết định
0`+7+B
0
T4PSvW]&++028HO*+888Fd0`F0`
+>F[-0+6+>+&85ZHO*+01•2FNFZ•0P+=<S78H
3000124e0<A+,N.BFACFA5
:Sl008N88.I0123+4,7<;OO*+8
88+ZO01,O*+888Wn<N+014=FI<•B24e0
<A+„…5
93h'+-&B(-*?-*[I-*H()-+(e(i&RC^-+&(RD-R+D5^=jg
B!$C D
+78+6+Dsp&O+M>+&>,FEFO*++&B<+V
8.BFA,_<]5 +0@3+78+&,,A0+'O&00+FA
+&O*++78[800+FA0PS}+>O*+4"VA4P5R04"VA
0P+=FO*++&,CO*+,_<]5
4"VA4,FO*++&,CH<+Sv4&00SF74=+'O
&4e0+&0@01+&,C4PHS4PSS,N004e0+&0@00+FA
0PS}4=+'O&00+FA8![8C+,N01+&,C4P5&<+H
O6'0@<+,N<+^>&0b>5ZH<+++78
+6+F<+4[08++&=TOc04*05:l,N<+^>&0
b>H01+&,C4,F+78[800AO5,N<+++78
+6+H01+&,C4,0P+=FO*+01,_0PBu5A+lSv+'O
0081+-0Bu0@01+&,C4=+'O&4e0+&0@P5
POQR) >12
Nhận kết quả
Đánh giá kết quả
Tốt ?
Dừng lại
Đúng
Công thức hoá lại
Sai
Hình 2. Mô hình thu thập thông @n chuẩn
2
0
1
1
2
Bài tập lớn môn Xử lý ngôn ngữ tự nhiên
$A0++78+6+0P+=4[04u0<C+0cO*+FI+6+
,-./V=,_<]H']H1O+555ZHT410j+0Y4"
07843,A0++78,_<]<T,_<]FO*+FI+6+O8+c0
+d0A,Vt+7+WJF\4]5RP+=COI&x00Vt+7+
0%0P+=4[08./0++78+6+48+A5
R0Vt+7+++78+6+0P+=4[00&+FI(
? R0Vt+7+09
{ }
0I
% H555HH
=
HT4P+
FO*++7+BWC+A+&+FA%
I
5)*+
+&,C4[0<=.<xO*+<=+c0F00@00+7+B<;O00
++JKGHLMH,L5
$-./,N+&,C(
RSNTU#NUVOOVWB
POQR) >14
2
0
1
1
2
Bài tập lớn môn Xử lý ngôn ngữ tự nhiên
:&O*++7800+FA,O*++&,C
{ }
U
### H555HH
,O*++&,C
|H0j+4+-+&0@OSSKNR#
O5:=74[000+FA
FZ2H0j+0Sf8W3800+FA]O.0@OSS,<•4+C+0]
00+FAc,NOSS•O*+6[240+&N05•
0P+=4[04uO*++&OSSKe0FO*+-+&SS,N
O*+,_<]4P50+c0FNC+0@Vt+7+F+'O4[0000`
+&•0f+8![85
:=+d0AO6'+'OV3O F>H0j+0P+=SJ./O*+SlVt
+7+A2]5ZH00+7++4PV64[04"078+&<Z
0c5
POQR) >15
2
0
1
0@+aV5
#SNJ
J
(((J
U
O
$-./HV81+-0+FA#
,#
F<Z0cHFZ2
43<A414H+0P,>0?+4[0I+&Z4;+?0"S(
POQR) >16
2
0
1
1
2
Bài tập lớn môn Xử lý ngôn ngữ tự nhiên
(0.6, 0.2)
Hình 3. Đồ thị biểu diễn các vec-tơ của bài báo D
1
và D
2
1.
0
1.
0
Magê
Đau đầu
D1
D2
query
Hình 4: Đồ thị biểu diễn quan hệ giữa truy vấn (query) và các tài liệu
D1, D2
2
0
1
1
2
55 +=
„…
WF>,+00F>+
F=
%d
T4Pd
FW0SC+WC+A+7+BX
+&+FA51Sl+&%d4[0
+-+<x888]+-0,NV]_WC+AX
+&+FA5
(Q( (( ( _6;A+ND8E-O
00+&`Sl00+7+B4[0W04H0j+0O*+OSf8W38
4=44*+4;B00,>0?++&,C,+FA5)*+Sl8n844*
+4;4[0+=A.N415g4P|,GFF[+F00+78+7+B+&
+&,C,+&,_<](
#R ∩
06+c04]C+
%R
#R
+
∩
ASl0@G0>
)*+44*+4;+6+XH4[0<3+4340S
„…HW04P0B,>0?++FA,,>0?++&,C<T8n8+-+O*+
V3+2]*+I5:e0<A+H4+X4[0+-,N4*.0@,>0?+5:*
+4;4[0W04+>06+c0.N41„…(
∑ ∑
∑
= =
=
==
J
J
R#
R#
R#D
5
5
EHD
•]SJ0]+&,C,+FA4[009<T4*.0@0jH06+c0
Sv+&TZ4](
2
0
1
1
2
Bài tập lớn môn Xử lý ngôn ngữ tự nhiên
:=N0F[A2]0@8n809H0j+SJ./<'8Sl
F[00+7+B+&O*++FA81Sl095$N00+&X[8
+FAf+'8n8+-WC8WYV64[00-W0H+ZVt+7+0%
0PO*+Sl4=OS(
? ’T0@V-0+N0+FA+&TZV60P\u,N<C+0c
FI095
? :*8c0+I8+-+•&C+"S,N00Vt+7++&N041
? RP+=+-++&N0
,7H4*+4;0P+=4[0+d0A<T06+c0S(
G+&>OVS
EHD
∑
=
=
+h2+0@OI?&D>&F>+m&VE5
POQR) >20
2
0
1
1
2
Bài tập lớn môn Xử lý ngôn ngữ tự nhiên
$N,A0SJ./OI?&H0j+0P+=<=.O*+8+[
+&+&+c0+&Fu,d00@<+H,0P+=4[0SJ./+06+&
A+l++78+6+5
(Q( ( ((/,8;PE
:=0P+==FO+3OI?&0P+=8./0WJF\++78
+6+H0j+Sv4uO*+SlVAO4[0SJ./+&F\+Z+
OI?&5
w1.d00Vl0@O6'+-+0OI?&+004,
`FiNME%MOO&C+"004e0+-0@&?&S`0„…H
P4jF00j+OI4[0O68•+>00?&0@4*,7+5
g00j+OIT'+=A008n8+F0KG5:&0@j+
OISvD3004,4"D5P4[0+d0A<T8n8SS,N
+&•NOOO`4&4"0P5=ZF,A0+d0A8n8F0
1
2
Bài tập lớn môn Xử lý ngôn ngữ tự nhiên
(Q( ( (( KE-6:28*
)6'4[04"WC+FO*+O6'<FN8(
? N800+7+B+&,CD00j+OI4,EQ|F>&
? N800+FAD00j+OI4&EQGF>&
? N8000YO/0D00j+9EQF>&
&'H0j+0P00V\AS(
?
+
?0YO/0+7+B
?
G
?+FA
?
|
?+7+B+&,C0@X.!
?
8
ˆ
?+&`SlV3+lB_0@OI+7+B,O*+_0@
OI+FA
?
2
?+&`SlFZV3+B+7+B0@+&,C,+7+B+
2
0
1
1
2
Bài tập lớn môn Xử lý ngôn ngữ tự nhiên
N8+7+B+&,C<=.00Z0X.!5){O*+j+FO*+
+7+B+&+&,C5N8+FA<=.+7800+FA5){j+203
43O*++FA5R0j++&FN80P00FZV3+0"0P+&`SlH+=A
Sd+4;B00+FA5•+&+4;4[0+-+<T+&`Sl
Bu0@00+7+B+&O{+FA5N800+7+BFFN84*5){
j+<=.O*++7+B4[040YO/05R0FZV3+0P+&B00j+
F00V3+FlB00+7+B+&8+&,C5
R0FZV3+0P+&`SlB00_+*0FN8+7+B0P0Y.y,00
_+*0FN8+7+B
I
<=.V]_\u0@+7+B
+&
+FA#
I
5ZV3+0P+&`Sl,
4u4*2+&`0@+7+B
+&
+<*+h[800+FA5R0+&VT40@00+&`Sl0P+=4[0
+&`yZe0,N<C+0c00W0453+&yZ4[0SJ
./H0P+=+Sv8]4lOe+,N00,C4"S(
2
0
1
1
2
Bài tập lớn môn Xử lý ngôn ngữ tự nhiên
? ++78+6+
? `0
++78+6+<f+4VX.!J0A+lO*+Z0
D+X4[0,3+.N.I6B+dZE5ŠZ04[081+-0,
+cFN8|Sv4[0W1.d5){_+&FN8|4[0FZV3+,NO*+
_+&FN8,N0!+7+B5ZV3+Sv<f+44[0V-0I+.`0
+>00OlFZV3+,
5){_0@FN87O*++-A+aFN8|Sv+-+
&x4*V-0I+,S4P+&"P+NOI5
&+&X[84P0PV]_0P+=W]&(
? &"F+-A+N_V0+aD+d4*+-+FIZ0E
? &"FP+N_4P+aG
&+&X[8+cHO{_0@FN8GSv+-O*++&V-0I+
8]WI4*+4;BZ0,+FA5R0+FA++784[0Sf8W38
POQR) >24
2
0
1
1
2
Bài tập lớn môn Xử lý ngôn ngữ tự nhiên
h3BP)+a+I*BIBIMk()P*&+D-(^CD;DDE-+(+()
h324P-RW()Pc^5_(`a(-+A()+,-
3$A+FO*+&A;„…H4e04=O<2++<*4e0
+&+3$A+,"Oe+B1OHBu,B885G4PH0j+8]+3
Z0c4e04=O0@+3$A+4=0P+=0P4[0BN
Z0c0/+=,",_<]+3$A+5
B1!72?:>O
B+>7P78:>
&+3$A+H0%+&00,_<]+3$A+H+0P+=+C1
FO*++8V2+&`5&V-88HO{+34c4*0F78H,+0P
+=8+A4[0001+&0]+3P0%+&,_<]„…5
Q((((1EL&
B1O0-FOe+1O0@6B5IS+FI8]Z0cV-
0I0@6B+3$A+5:PF,'+&Z+d0+3H00c./FZ243