A User Behavior Analysis Method Based On Big Data
A User Behavior Analysis Method Based On Big Data
$EVWUDFW² :LWK WKH GHYHORSLQJ SURJUHVV RI WKH ,QWHUQHW DQG SXEOLVKHG RQ 7LDQFKL ZKLFK LV WKH ,7 GHYHORSHU QHWZRUN
WKH SDQGHPLF GHWHULRUDWLRQ PRUH SHRSOH GHFLGH WR VKRS RQOLQH SODWIRUP RI 7DREDR %\ RIIHULQJ && DQG %& VKRSSLQJ
7KH RQOLQH H&RPPHUFH EXVLQHVV LV TXLFNO\ H[SDQGLQJ 7DREDR LV WKH ODUJHVW H&RPPHUFH UHWDLO SODWIRUP LQ &KLQD >@
8QGHUVWDQGLQJ XVHUV¶ SXUFKDVH EHKDYLRU LV D NH\ FRPSRQHQW WR 7KH GDWD ZDV DERXW RQH PLOOLRQ XVHUV ZKR KDG EHKDYLRUV RI
JURZLQJ VDOHV DQG LPSURYLQJ WKH ZHEVLWH¶V GHVLJQ DQG IORZ :H SDJH YLHZV SXUFKDVH VKRSSLQJ WR FDUW DQG DGGLQJ WR
XVH 8VHU %HKDYLRU 'DWD IURP 7LDQFKL RI $OLEDED &ORXG 7KH IDYRULWH GXULQJ 1RYHPEHU WR 'HFHPEHU 7KHUH
GDWD ZHUH UDQGRPO\ VHOHFWHG DERXW RQH PLOOLRQ XVHUV ZKR KDYH DUH D WRWDO RI WHQ PLOOLRQ UHFRUGV ,W FRQWDLQV ILYH FROXPQV
EHKDYLRUV RI SDJH YLHZ SXUFKDVH VKRSSLQJ WR FDUW DQG DGGLQJ 8VHU,' ,WHP,' &DWHJRU\,' %HKDYLRU 7\SH DQG
WR IDYRULWH GXULQJ 1RYHPEHU WR 'HFHPEHU 7KHUH 7LPHVWDPS
DUH D WRWDO RI WHQ PLOOLRQ UHFRUGV :H XVHG IRXU GLPHQVLRQV WR 8VHU,' LV LQ VHULDOL]HG LQWHJHU IRUPDW ZKLFK HDFK
DSSURDFK WKLV SUREOHP WLPH LWHP FRQYHUVLRQ UDWH IURP UHSUHVHQWV D XVHU
)XQQHO $QDO\VLV DQG YDOXDEOH FXVWRPHUV IURP 5)0 0RGHO ,WHP,' LV LQ VHULDOL]HG LQWHJHU IRUPDW ZKLFK HDFK
)URP WKH WLPH GLPHQVLRQ XVHUV DUH PRUH OLNHO\ WR YLVLW SDJHV UHSUHVHQWV D XQTLXH LWHP
GXULQJ WKH ZHHNHQGV IURP DP WR SP )URP WKH LWHP &DWHJRU\ ,' LV LQ VHULDOL]HG LQWHJHU IRUPDW ZKLFK
GLPHQVLRQ WKH WRS WHQ LWHPV ZHUH SXUFKDVHG RYHU RQH KXQGUHG
HDFK UHSUHVHQWV WKH LWHPV EHORQJHG FDWHJRU\
WLPHV GXULQJ WKLV WLPH IUDPH 7KH QXPEHU LV RXWVWDQGLQJ EDFN
LQ )RU WKH FRQYHUVLRQ UDWH IURP )XQQHO $QDO\VLV DGGLQJ
%HKDYLRU W\SH LV D VWULQJ WKDW FRQVLVWV WKH VHW RI SY
LWHPV LQWR WKH FDUW KDV D KLJKHU FRQYHUVLRQ UDWH FRPSDUHG WR EX\ FDUW DQG IDY
PDUNLQJ LWHPV DV IDYRULWHV )RU WKH FXVWRPHU GLPHQVLRQ ZH D SY UHSUHVHQWV SDJH YLHZ RI DQ LWHP V GHWDLO SDJH
XVHG DQ 5)0 UHFHQF\ IUHTXHQF\ PRQHWDU\ PRGHO WR LGHQWLI\ HTXLYDOHQW WR DQ LWHP FOLFN
YDOXDEOH XVHUV :H FDQ DGYLVH 7DREDR WR LPSURYH WKH IDYRULWH E EX\ UHSUHVHQWV SXUFKDVH DQ LWHP
LWHP UHPDLQGHU SURFHVV LQFUHDVH WKH H[SRVXUH RI KRW VHOOLQJ F FDUW UHSUHVHQWV DGG DQ LWHP WR VKRSSLQJ FDUW
LWHPV GXULQJ SHDN KRXUV DQG ZHHN RI WKH GD\ DQG RIIHU PRUH
SURPRWLRQVUHFRPPHQGDWLRQV WR YDOXDEOH FXVWRPHUV G IDY UHSUHVHQWV PDUN IDYRU RI DQ LWHP
7LPHVWDPS LV LQ VHULDO LQWHJHU IRUPDW ZKLFK HDFK
.H\ZRUGVH&RPPHUFH SHDN EX\LQJ WLPH KRW VHOOLQJ LWHPV UHSUHVHQWV WKH WLPHVWDPS RI WKH EHKDYLRU
SXUFKDVH FRQYHUVLRQ UDWH )XQQHO $QDO\VLV 5)0 0RGHO
% 'DWD (7/ 3URFHVV
, ,1752'8&7,21 )RU WKH VWRUDJH RI WKH RULJLQDO GRZQORDGHG GDWDVHW DQG LWV
:LWK WKH FRQWLQXHG JURZWK RI LQWHUQHW VKRSSLQJ IXHOHG HGLWHG QHZ YHUVLRQV ZH XVHG +DGRRS IRU VWRUDJH DQG
E\ WKH &29,' SDQGHPLF LQWHUQHW VKRSSLQJ FRQWLQXHV WR WUDQVIRUPDWLRQV YLD S\WKRQ ³+')6´ PRGXOH
DXJPHQW 6KRSSLQJ RQOLQH ULVHV WR PRUH WKDQ DIWHU WKH
,,, )285 $1$/<7,&$/ ',0(16,216 ',6&866,21
&29,' SDQGHPLF EHJDQ >@ 6LQFH HFRPPHUFH LV
TXLFNO\ H[SDQGLQJ XQGHUVWDQGLQJ FRQVXPHU SXUFKDVH $V SUHYLRXVO\ PHQWLRQHG ZH DSSURDFK E\ XVLQJ IRXU
EHKDYLRU KDV EHFRPH DV DQ SUDFWLFDO DQDO\WLFDO WRRO IRU GLPHQVLRQV 8VHU 3XUFKDVH ,WHP 3XUFKDVH )XQQHO $QDO\VLV
JURZLQJ VDOHV DQG LPSURYLQJ WKH ZHEVLWH V XVHU H[SHULHQFH DQG 9DOXDEOH 8VHUV
>@
$ 8VHU 3XUFKDVH
:H XVH WKH GDWDVHW ³8VHU %HKDYLRU 'DWD RQ 7DREDR IRU
5HFRPPHQGDWLRQ´ WR LGHQWLI\ FUXFLDO IDFWRUV WKDW DIIHFW :H XVHG 39 SDJH YLHZ 89 XVHU YLHZ ERXQFH UDWHV
FRQVXPHU EHKDYLRU ZH XVH IRXU GLPHQVLRQV 8VHU 3XUFKDVH DQG RWKHU LQGH[HV WR DQDO\]H XVHU DFWLYH WLPH XQGHUVWDQGLQJ
,WHP 3XUFKDVH )XQQHO $QDO\VLV DQG 9DOXDEOH 8VHUV XVHU¶V EHKDYLRUV
6HFWLRQ ,, LV DERXW WKH GDWD VRXUFH DQG GDWD (7/ SURFHVV % ,WHP 3XUFKDVH
6HFWLRQ ,,, LV DERXW WKH GLVFXVVLRQ RQ IRXU DQDO\WLFDO
GLPHQVLRQV 6HFWLRQ ,9 LV DERXW GDWD FOHDQLQJ DQG DQDO\]LQJ :H DSSURDFK WKLV GLPHQVLRQ E\ QXPEHU RI WUDQVDFWLRQV
WKH GDWD /DVWO\ 6HFWLRQ 9 LV DERXW FRQFOXVLRQ DQG DYHUDJH SXUFKDVHU SHU XVHU PXOWLSOH SXUFKDVH UDWH DQG RWKHU
GLVFXVVLRQ LQGH[HV WR H[SORUH XVHUV SXUFKDVH SUHIHUHQFH DQG XQGHUVWDQG
SXUFKDVH SDWWHUQ
,, '$7$ 6285&( $1' '$7$ (7/ 352&(66
& )XQQHO $QDO\VLV
$ 'DWD 6RXUFH %\ DQDO\]LQJ SDJH YLHZ FDUW DQG IDYRULWH WR EX\
)RU WKLV XVHU EHKDYLRU DQDO\VLV ZH XVH WKH GDWDVHW ³8VHU WUDQVLWLRQ UDWH ZH ILQG ZKLFK OD\HUV QHHG WR EH LPSURYHG
%HKDYLRU 'DWD RQ 7DREDR IRU 5HFRPPHQGDWLRQ´ >@
)LJ 6DPSOH RI ILUVW UHFRUGV DIWHU GDWD FOHDQLQJ SURFHVV G %RXQFH 5DWH 8VHUV 7RRN 1R $FWLRQV $IWHU 3DJH
9LHZ
7KHUH DUH UHFRUGV DIWHU FOHDQLQJ VKRZQ LQ %RXQFH 5DWH LV DURXQG )LJ 7KLV UHVXOW LV QRW
)LJ /DUJH HQRXJK IRU XV WR DQDO\]H EDG FRPSDULQJ WR RQ DYHUDJH RI 7DREDR RIILFLDOO\
SXEOLVKHG JHQHUDO %RXQFH 5DWH 7KLV PHDQV WKH LWHP SDJH
FRXOG DWWUDFW HQRXJK XVHUV IRU QH[W VWHSV RI HLWKHU EX\
IDYRULWH RU FDUW
% 'DWD $QDO\VLV
)LJ %RXQFH UDWH FDOFXODWLRQ
8VHU 3XUFKDVH
D &RXQW RI 3DJH 9LHZ H 'DLO\ 3DJH 9LHZ DQG 8QLTXH 8VHUV
:LWKLQ GD\V ZH KDYH D WRWDO RI SDJH YLHZV )URP )LJ ZH FDQ VHH WKDW WKHUH DUH UHODWLYHO\ PRUH
)LJ SDJH YLHZ FRXQWV EHWZHHQ RQ WR DQG WR
7KH UHDVRQ PLJKW EH WKHVH GD\V DUH )ULGD\V DQG
ZHHNHQGV WKHUHIRUH SHRSOH KDYH PRUH WLPH WR YLVLW 7DREDR
IRU VKRSSLQJ
7KHUH DUH VLPLODU XQLTXH XVHUV UHJDUGOHVV WKH WLPH RI WKH
ZHHN H[FHSW WR 7KH FRXQW RI XQLTXH XVHUV UHDFKHV
WR PD[ DW 7KDW GDWH LV 6DWXUGD\ 7KH UHDVRQ ZK\
KDV PRUH XQLTXH XVHUV WKDQ ODVW 6DWXUGD\ LV SUREDEO\
)LJ 7RWDO SDJH YLHZV ZLWKLQ GD\V WKDW GXULQJ ZHHNHQG 7DREDR KDV VRPH VLWHZLVH
SURPRWLRQV DWWUDFWV PRUH XVHUV WR YLVLW RQ WKH ZHEVLWH
E &RXQW RI 8QLTXH 8VHUV
:H KDYH XQLTXH XVHUV )LJ
492
Authorized licensed use limited to: Genba Sopanrao Moze College Of Engineering. Downloaded on September 28,2024 at 07:43:10 UTC from IEEE Xplore. Restrictions apply.
E $YHUDJH 3XUFKDVH 3HU 8VHU E\ 'DWH
$YHUDJH SXUFKDVH SHU XVHU E\ GDWH LV DERXW KRZHYHU
RQ LW LV RQO\ DQG UHDFKHV WR WKH ORZHVW RI
)LJ ,W VHHPV OLNH ZLWK KLJKHVW SXUFKDVH DQG
FXVWRPHUV WKH DYHUDJH SXUFKDVH SHU XVHU PLJKW QRW EH WKH
KLJKHVW DV ZHOO
,WHP 3XUFKDVH
D 'DLO\ 6XFFHVVIXO 3XUFKDVH
6WDUWLQJ IURP WKH VXFFHVVIXO SXUFKDVH FRXQW LV
VLPLODU IRU HDFK GD\ XQWLO )LJ 2Q 7KH
VXFFHVVIXO SXUFKDVH LV ZLWK XQLTXH XVHUV
SDUWLFLSDWHG
)LJ 7RS XVHUV
G 5HSXUFKDVH 5DWH
5HSXUFKDVH UDWH LV DERXW )LJ 7KLV PHDQV
DOPRVW RXW RI HYHU\ XVHUV PDNH VHFRQG RU PRUH
SXUFKDVHV
493
Authorized licensed use limited to: Genba Sopanrao Moze College Of Engineering. Downloaded on September 28,2024 at 07:43:10 UTC from IEEE Xplore. Restrictions apply.
H +RW 6HOOLQJ ,WHPV D )DYRULWH 72 %X\ &RQYHUVLRQ 5DWH
7KH LWHP JRW SXUFKDVHG PRVW LV LWHP ZLWK 7KH FRQYHUVLRQ UDWH IURP IDY WR EX\ LV DERXW
WLPHV )LJ
6HFRQG KLJKHVW LV LWHP ZLWK WLPHV
7KLUG KLJKHVW LV LWHP ZLWK WLPHV
)LJ OLVWV WRS KRW VHOOLQJ LWHPV
494
Authorized licensed use limited to: Genba Sopanrao Moze College Of Engineering. Downloaded on September 28,2024 at 07:43:10 UTC from IEEE Xplore. Restrictions apply.
)UHT 5DQN ! DQG
,PSRUWDQW .HHS 8VHUV
%X\ 5DQN
)UHT 5DQN DQG
,PSRUWDQW 'HYHORSLQJ 8VHUV
%X\ 5DQN!
)UHT 5DQN DQG
5HJXODU 8VHUV
%X\ 5DQN
E )UHTXHQF\
)RU )UHTXHQF\ IURP WKH HDUOLHU XVHU PRVW SXUFKDVHG
DQDO\VLV ZH JHW DV TXDQWLOH ,Q WKLV FDVH ZH XVH
DV WKH UHIHUHQFH UDQN WR FODVVLI\ DV OHYHOV
x DV
x DV
x DV
x DV
x ! DV )LJ 8QLTXH FXVWRPHUV LQ GLIIHUHQW FODVVHV
)LJ VKRZV VDPSOHV RI UDQNLQJ FODVVLILFDWLRQ
)RU ,PSRUWDQW 9DOXDEOH 8VHUV WKLV JURXS LV WKH EHVW :H
VKRXOG VSHQG PRUH WLPH DQG IRFXV RQ WKLV JURXS WU\ WKH EHVW
WR DFFRPPRGDWH WKRVH XVHUV 7KHVH XVHUV DUH RXU WRS SULRULW\
)RU ,PSRUWDQW .HHS 8VHUV WKLV JURXS KDV EHHQ SXUFKDVHG
UHFHQWO\ EXW QRW WRR PDQ\ WLPHV :H VKRXOG JHQHUDWH PRUH
VSHFLDO SURPRWLRQV IRU WKLV JURXS WR WU\ WR EULQJ XS WKH
SXUFKDVH IUHTXHQF\
)RU ,PSRUWDQW 'HYHORSLQJ 8VHUV DOWKRXJK WKH\ KDYH QRW
SXUFKDVHG UHFHQWO\ WKH\ SXUFKDVHG PXOWLSOH WLPHV EHIRUH
:H VKRXOG EXLOG XS D UHDFK FKDQQHO IRU WKLV JURXS
)RU 5HJXODU 8VHUV WKLV JURXS SXUFKDVHV QHLWKHU
IUHTXHQWO\ QRU UHFHQWO\ 7KH SODWIRUP ZRXOG HDVLO\ ORVH WKLV
JURXS WKHUHIRUH WKH SODWIRUP VKRXOG RIIHU PRUH FRXSRQV DQG
SXVKQRWLILFDWLRQV WR HYRNH WKH ZLOOLQJQHVV WR SXUFKDVH
)LJ &DUW WR EX\ FRQYHUVLRQ UDWH
9 &21&/86,21 $1' ',6&866,216
7KH SDJH YLHZV DQG VDOHV RQ DQG DUH
UHODWLYHO\ KLJKHU WKDQ RWKHU GDWHV 7KLV PLJKW EH EHFDXVH WKH
)LQDOO\ ZH FRPSDUH WKH UHVXOW IURP 5 DQG ) E\ XVLQJ VLWHZLVH SURPRWLRQ
PHDQ DQG WKHQ FODVVLI\ XVHUV E\ WKH PHDQV )RU 5HFHQF\ 7KH SHDN DFWLYH WLPH RI WKH ZHHN LV WKH ZHHNHQG ZH
5DQN ZH KDYH PHDQ HTXDO WR )RU )UHTXHQF\ 5DQN KDYH PRUH XVHUV DQG SDJH YLHZV EHKDYLRUV
ZH KDYH PHDQ HTXDO WR )LJ 7KH SHDN WLPH RI WKH GD\ LV DW DQG SP
7KH SODWIRUP FRXOG FUHDWH VRFLDO PHGLD JURXSV IRU WKH
WRS KRW VHOOLQJ LWHPV WR FROOHFW OR\DO XVHUV IRU WKRVH LWHPV
7KH SODWIRUP VKRXOG IRFXV RQ WKH WRS SXUFKDVH XVHUV
HVSHFLDOO\ 8VHU ,' ZKR SXUFKDVHV PRUH WKDQ
WLPHV RYHU GD\V %XLOGLQJ VSHFLDO XVHU GDWDEDVHV DQG
WU\LQJ WR IXOO\ XQGHUVWDQG WRS SXUFKDVH XVHUV ZRXOG KHOS WKH
SODWIRUP JHQHUDWH PRUH VDOHV DQG SHUVRQDOL]HG VHUYLFH WR
WKRVH WRS SXUFKDVH XVHUV
2QO\ SDJH YLHZ WR EX\ FRQYHUVLRQ UDWH
5HFRPPHQG RIIHULQJ PRUH SURPRWLRQV LPSURYLQJ LWHP
)LJ 0HDQV RI 5HFHQF\ 5DQN DQG )UHTXHQF\ 5DQN
GHVFULSWLRQ SDJHV DQG VPDUWHU D FKHFNRXW SURFHVV WR
LQFUHDVH WKH FRQYHUVLRQ UDWH
8VHU FODVVLILFDWLRQ 7DEOH
&DUW WR %X\ FRQYHUVLRQ UDWH LV KLJKHU WKDQ )DYRULWH WR
7$%/( , 86(5 &/$66,),&$7,21 %X\ FRQYHUVLRQ UDWH )RU FDUW WR EX\ ZH UHFRPPHQG WKH
SODWIRUP WR VHQG UHPDLQGHUV WR LQGXFH XVHUV WR FKHFNRXW
)UHT 5DQN ! DQG
,PSRUWDQW 9DOXDEOH 8VHUV
%X\ 5DQN!
LWHPV LQ WKH FDUW )RU IDYRULWH WR EX\ ZH UHFRPPHQG WKH
495
Authorized licensed use limited to: Genba Sopanrao Moze College Of Engineering. Downloaded on September 28,2024 at 07:43:10 UTC from IEEE Xplore. Restrictions apply.
SODWIRUP WR JHQHUDWH PRUH GLVFRXQWV DQG UHPLQGLQJ IURP KWWSVPHGLXPGDWDGULYHQLQYHVWRUFRPZK\FRQVXPHU
QRWLILFDWLRQV WR WKRVH XVHUV EHKDYLRUDQDO\VLVLVVRUHOHYDQWWRWKHHFRPPHUFHEXVLQHVV
IFFDF
7KURXJK 5)0 PRGHO FODVVLILFDWLRQ WKHUH DUH PRUH
>@ 'DWDVHW7,$1&+, $OLEDED &ORXG QG IURP
XVHUV LQ ,PSRUWDQW .HHS 8VHUV DQG 5HJXODU ,W LV KWWSVWLDQFKLDOL\XQFRPGDWDVHWGDWD'HWDLO"GDWD,G XVHU,G
UHFRPPHQGHG WR FRQGXFW WDUJHWHG SUHFLVLRQ PDUNHWLQJ >@ $%1 1HZVZLUH QG 7DREDR 7DREDR _ &RPSDQ\ 6XPPDU\ _ IURP
EDVHG RQ WKHVH XVHU JURXSV KWWSZZZDEQQHZVZLUHQHWJIZFRPSDQLHVHQ7DREDR
>@ 6(*$/ 7 5 2 < $XJXVW )XQQHO $QDO\VLV +RZ WR XVH
5()(5(1&(6 LW WR ERRVW FRQYHUVLRQV DPS JURZWK +RWMDU EORJ )XQQHO $QDO\VLV
>@ (FROD / /X + DPS 5RKU & -XO\ 2QOLQH VKRSSLQJ +RZ WR 8VH ,W WR %RRVW &RQYHUVLRQV DPS *URZWK _ +RWMDU %ORJ
0DQ\ $PHULFDQV KDELWV KDYHQ W FKDQJHG GXULQJ &29,' 5$1' IURP KWWSVZZZKRWMDUFRPEORJIXQQHODQDO\VLVZKDWLVIXQQHO
&RUSRUDWLRQ IURP DQDO\VLV
KWWSVZZZUDQGRUJSXEVUHVHDUFKBUHSRUWV55$KWPO >@ 6HJDO 7 6HSWHPEHU 5HFHQF\ IUHTXHQF\ PRQHWDU\ YDOXH
>@ &RXQWDQWV -DQXDU\ :K\ FRQVXPHU EHKDYLRU DQDO\VLV LV VR 5)0 GHILQLWLRQ ,QYHVWRSHGLD IURP
UHOHYDQW WR WKH HFRPPHUFH EXVLQHVV" 0HGLXP 5HWULHYHG 2FWREHU KWWSVZZZLQYHVWRSHGLDFRPWHUPVUUIPUHFHQF\IUHTXHQF\
PRQHWDU\YDOXHDVS
496
Authorized licensed use limited to: Genba Sopanrao Moze College Of Engineering. Downloaded on September 28,2024 at 07:43:10 UTC from IEEE Xplore. Restrictions apply.