[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Tho^'ng ke^ truye^.n Kie^`u, Chinh Phu. & Cung Oa'n



Cha`o anh Trieu va` anh AiViet,

        Nhu+ va^.y anh va` to^i co' cu`ng chi' hu+o+'ng! :) 
Hay la` mi`nh ho+.p ta'c la`m the^m? Ba^y gio+` dde^? to^i 
ha^`u chuye^.n anh: 

>1. Ca? ba truye^.n du`ng nhie^`u va^`n ba(`ng ho+n va^`n tra('c.
>    DDie^`u na`y co' le~ vi` quy lua^.t tho+ ba('t buo^.c du`ng nhie^`u
>    va^`n ba(`ng nhie^`u ho+n va^`n tra('c. The^m mo^.t ly' do nu+~a la`
>    va^`n ba(`ng ddo.c nghe e^m tai ho+n va^`n tra('c.

        DDo^`ng y'.

>2. Truye^.n Kie^`u du`ng nhie^`u va^`n ba(`ng (60\%) ho+n Chinh Phu. va`
>   Cung Oa'n (57\%).
>    Ly' do co' the^? la` truye^.n Kie^`u vie^'t theo the^? tho+ lu.c ba't
>        B B T T B B
>        B B T T B B T B
>    ba('t buo^.c du`ng 9 va^`n ba(`ng (64\%) va` 5 va^`n tra('c (36\%).
>    Trong khi ddo' Chinh Phu. vie^'t theo the^? tho+ song tha^'t lu.c
>    ba't. Off-hand, to^i kho^ng co' sa'ch dde^? check quy lua^.t cu?a
>    hai ca^u 7-chu+~ cu?a tho+ song tha^'t lu.c ba't
>        . . . . . . .
>        . . . . . . .
>        B B T T B B
>        B B T T B B T B
>    nhu+ng ddie^`u na`y co' the^? kie^?m la.i de^~ da`ng. Ma(.t kha'c,
>    60\% cu?a Kie^`u va` 57\% cu?a Chinh Phu. va` Cung Oa'n cu~ng kho^ng
>    kha'c xa nhie^`u la('m.

	No'i ve^` ca'ch gieo va^`n thi` la.i la` mo^.t ddie^`u 
kha' thu' vi. . Theo nhu+ sa'ch nga`y xu+a ma` to^i ho.c 
ddu+o+.c thi` tho+ lu.c ba't thu+o+`ng theo ca'i pattern [nhu+ 
anh quote]

	b b t t b b   va`   b b t t b b t b

	Nhu+ng trong thu+.c te^', Truye^.n Kie^`u du`ng nhie^`u 
combination ho+n. Tha^.t ra, ca'i lua^.t tre^n dda^y chi? du`ng 
cho 24 la^`n [1.5%] trong Kie^`u ma` tho^i. Ti'nh ra cu. 
Nguye^~n dda~ du`ng dde^'n 156 combinations. Sau dda^y la` 89 
combinations thu+o+`ng du`ng nhie^`u nha^'t:
                                               Cumulative  Cumulative
        PATTERN           Frequency   Percent   Frequency    Percent
        -------------------------------------------------------------
        bbttbb tbttbbtb         26       1.6          26        1.6
        bbttbb tbbttbtb         25       1.5          51        3.1
        tbttbb tbbttbbb         25       1.5          76        4.7
        tbttbb tbbttbtb         25       1.5         101        6.2
     ** bbttbb bbttbbtb         24       1.5         125        7.7
        bbttbb tbtttbbb         23       1.4         148        9.1
        bbttbb bbtttbtb         22       1.4         170       10.4
        tbbtbb tbbttbbb         21       1.3         191       11.7
        tbbtbb tbttbbtb         21       1.3         212       13.0
        tbttbb bbtttbtb         21       1.3         233       14.3
        tbbttb tbbttbbb         20       1.2         253       15.6
        tbttbb bbttbbtb         20       1.2         273       16.8
        tbbttb bbttbbtb         19       1.2         292       17.9
        tbbttb tbttbbtb         19       1.2         311       19.1
        tbttbb bbbttbtb         19       1.2         330       20.3
        tbttbb tbtttbtb         19       1.2         349       21.5
        bbttbb bbbtbbtb         18       1.1         367       22.6
        bbttbb tbbtbbtb         18       1.1         385       23.7
        bbttbb tbtttbtb         18       1.1         403       24.8
        tbbttb bbtttbbb         18       1.1         421       25.9
        tbbttb tbbttbtb         18       1.1         439       27.0
        tbttbb bbbtbbbb         18       1.1         457       28.1
        tbttbb tbtttbbb         18       1.1         475       29.2
        bbbtbb tbbtbbtb         17       1.0         492       30.2
        bbttbb tbbttbbb         17       1.0         509       31.3
        tbttbb tbttbbtb         17       1.0         526       32.3
        bbttbb bbbttbbb         16       1.0         542       33.3
        bbttbb bbttbbbb         16       1.0         558       34.3
        bbtttb tbbttbtb         16       1.0         574       35.3
        tbbtbb bbbttbtb         16       1.0         590       36.3
        tbbttb tbtttbtb         16       1.0         606       37.2
        bbbtbb bbbtbbbb         15       0.9         621       38.2
        bbbttb bbbttbtb         15       0.9         636       39.1
        tbbtbb tbbttbtb         15       0.9         651       40.0
        tbbtbb tbtttbtb         15       0.9         666       40.9
        tbbttb bbtttbtb         15       0.9         681       41.9
        tbbttb tbbtbbbb         15       0.9         696       42.8
        tbbttb tbbtbbtb         15       0.9         711       43.7
        tbttbb tbbtbbtb         15       0.9         726       44.6
        tbtttb tbttbbtb         15       0.9         741       45.5
        bbbtbb tbbttbtb         14       0.9         755       46.4
        bbbtbb tbttbbtb         14       0.9         769       47.3
        bbbtbb tbtttbbb         14       0.9         783       48.1
        bbbttb tbtttbtb         14       0.9         797       49.0
        bbttbb bbbttbtb         14       0.9         811       49.8
        tbbtbb bbtttbtb         14       0.9         825       50.7
	......

	To^i co`n chia Kie^`u ra 30 ddoa.n [nhu+ trong VHVN]. 
Mo^~i ddoa.n to^i ti'nh ty? le^. [%] cu?a thanh ba(`ng, 
ro^`i ve~ tre^n mo^.t bie^?u ddo^` nhu+ sau:

                 
 % thanh ba(`ng
     |
62.5 +
     |         *               *             *     *         *
     | *     *       *     * *                       *     *   *
60.0 +     *                         *           *       *
     |   *       * *                           *                 *
     |                 *           *     *             *
57.5 +                           *     *
     |                   *                 *
     |
55.0 +
     --+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-  
       1 2 3 4 5 6 7 8 9 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 3  
                         0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0  

					Doa.n [1 to+'i 30]

	Anh tha^'y sao ? Rie^ng to^i thi` tha^'y co' mo^.t 
pattern le^n-xuo^'ng gio^'ng nhu+ sine function trong ty? 
le^. du`ng thanh ba(`ng ro~ ra`ng theo ddoa.n tho+ . Mai 
mo^'t se~ ti`m xem ca'i function na`y la` gi` ? 

>3. To^i kho^ng hie^?u ro~ ca'ch ti'nh cu?a Ba?ng 2 khi anh so sa'nh
>    frequency cu?a truye^.n va` random choice of words. Thi' du., Kie^`u
>    co' N=6232 chu+~ da^'u sa('c trong to^?ng so^' 22,778 chu+~. Anh
>    randomize ra sao ma` tha`nh expected frequency F=6070? To^i thu+.c
>    ti`nh kho^ng ro~. Ne^'u anh clarify the^m ddu+o+.c thi` to^'t la('m.

	A`! ca'i na`y thi` to^i ti'nh theo Chi-Square 
analysis ddo' ma`. 


>4. Typo (minor) trong ca^u ``... Ngu+o+.c la.i, CONK va` CONk du`ng
>    chu+~ da^'u huye^`n i't ho+n la` expected....'' Cha('c y' anh muo^'n
>    no'i CONK va` CPNK?

	Merci beaucoup anh Trieu. DDa^y la` su+. giu'p ddo+? 
ma` to^i ra^'t ca^`n. 

>
>To^i co' mo^.t va`i y' kie^'n dde^? la`m the^m nhu+ sau:
>
>5. Ne^'u co' the^? thi` mi`nh du`ng testing hypothesis dde^? back-up
>    nhu+~ng gia? thuye^'t va` ke^'t lua^.n cu?a mi`nh. Tuy nhie^n
>    ddie^`u na`y ddo`i ho?i mi`nh pha?i cho.n underlying distribution
>    cu?a chu+~ trong ca'c truye^.n na`y\.
>
>6. Mo^.t thi' nghie^.m interesting la` mi`nh ty? du. Kie^`u ba^y gio+`
>    vie^'t theo the^? tho+ song tha^'t lu.c ba't gio^'ng nhu+ Chinh Phu.
>    (hay la` ty? du. Chinh Phu. vie^'t theo the^? tho+ lu.c ba't gio^'ng
>    nhu+ Kie^`u). What would happen? Truye^.n co' vie^'t "hay" ho+n
>    kho^ng?

	Y' cu?a anh ra^'t hay. Tha^.t ra ca'c pha^n ti'ch ma` 
to^i tri`nh ba`y chi? la` pha^`n dda^`u tho^i. Kie^?u nhu+ 
la` exploratory, chu+' chu+a ddi va`o modelling gi` ca?. 
To^i cu~ng nghi~ tu+o+ng tu+. nhu+ anh ve^` point 5, nhu+ng 
chu+a nghi~ dde^'n point 6. Theo anh thi` to^i ne^n 
approach point 6 nhu+ the^' na`o ? 

	Ho^m na`y ra?nh to^i se~ ba`n the^m .

	Tua^'n