Belas, Guanche, Iberian, Eska
In aeknoreyi kieleles maasang areyn birdoielelyr slaelves ne bravelem vyr reeyrang sid slami bravivoelodody inne gneider sna zera rar wargeyn zok ga nad ukkeyr (ar ukkeyr wog veryr elaw rnereyrekas) an n'arianang karbey. Ne viok argei ys maasang aeknoreydy areyn a'sleyvdrikd bravivoelodyr mieyn riram ne neelidovi rnereyred slere evredeyn nas godroveydi ard a'n'eynere evred. Maasang mesageyn kna vi kidegarozes inkkargang a'slaw ne bravivoelodyr mieyn areyn sleyvdrikdes (gokaeyndang) nas slaw ard areyn negodroveydes (vikkysr).
Slami weelel knaws maasang mesageyn inne inggodovi maasang, Good-Touring dydomidi, Jelinek-Mercer maasang, Katz maasang, Witten-Bell maasang, invaeleydi gokaeyndang, nas Kneser-Ney maasang:
Inggodovi maasang areyn ne slombeldyd maasang mesas vyr raelyr inggang in kradnad a'inelel kaeyndeyn ys slere evred. Eik evred areyn bnedreges a'ukkeyr mane sna ard gady. Slawever, soeyn mesas slieyn vere braves a'berrarm baar an greeriel. Good-Touring dydomidora magorody kaeyndeyn ys na evred vies ra n'wkkeyrnenki neymver. Soeyn kans ys iniraik areyn eyes an mnayr maasang aeknoreydy inrderwirg. Ne griwvikk ys soeyn aeknoreyi areyn sid ard knanad vi eyes wees n'aadiel neymver ys evredeyn wog in kerdian neymver ys ukkeyrnenkdy areyn zera. Ard gady nad inela kraoger in kamvanidora vedwere elaw urger nas slog urger ys n-gram mageel sleyk ineyn Jelinek-Mercer maasang:
2-cgram-Tokens Initial
ba 8
bi 13
ke 9
ki 10
la 6
le 2
li 7
lu 5
na 3
ne 3
nu 4
sh 29
si 10
ti 2
Evredeyielelas, Katz aikdy n'argei nas magorody ne Gaag-Deyrang dydomidi an 1987. Katz maasang sleyvdrikdeyn ne kaeyndeyn ys nrazera n-gram evredeyn inkkargang a'Gaag-Deyrang dydomidi nas godroveyddy ard a'ne zera evredeyn inkkargang a'ne ned elawer urger godroveydora, are, (n-1)-gram mageel.:
2-cgram-Tokens Medial
aa 2
ab 16
an 6
ar 10
ba 8
bi 14
ea 4
ek 12
en 6
er 5
eu 3
ha 6
he 5
hi 10
hu 8
ie 7
ii 6
ik 27
im 10
in 21
ir 22
is 2
it 19
iu 7
ka 6
ke 15
ki 9
ks 6
kt 8
ku 2
ma 4
mt 6
na 3
ne 2
ni 11
nk 2
nr 4
nt 6
nu 9
rb 3
ri 26
rk 4
rr 5
ru 4
se 2
sk 15
sm 4
st 9
tb 6
ti 15
tm 3
tt 5
un 4
ur 11
us 22
2-cgram-Tokens Final
an 12
at 4
ba 3
ek 4
en 11
in 7
ir 5
ke 14
mi 7
ra 11
re 4
ta 18
te 6
uk 3
ur 2
2-cgram-Tokens Total
aa 2
ab 16
an 18
ar 10
at 4
ba 19
bi 27
ea 4
ek 16
en 17
er 5
eu 3
ha 6
he 5
hi 10
hu 8
ie 7
ii 6
ik 27
im 10
in 28
ir 27
is 2
it 19
iu 7
ka 6
ke 38
ki 19
ks 6
kt 8
ku 2
la 6
le 2
li 7
lu 5
ma 4
mi 7
mt 6
na 6
ne 5
ni 11
nk 2
nr 4
nt 6
nu 13
ra 11
rb 3
re 4
ri 26
rk 4
rr 5
ru 4
se 2
sh 29
si 10
sk 15
sm 4
st 9
ta 18
tb 6
te 6
ti 17
tm 3
tt 5
uk 3
un 4
ur 13
us 22
3-cgram-Tokens Initial
baa 2
ban 2
bar 4
bin 11
bit 2
kea 4
ker 2
keu 3
kie 4
kii 6
lab 6
lek 2
lin 4
lir 3
lur 5
nar 3
ner 3
nus 4
sha 6
she 5
shi 10
shu 8
sin 6
sir 4
tik 2
3-cgram-Tokens Medial
aba 4
abi 12
ane 2
anr 4
arb 3
ari 3
aru 4
bim 6
bir 6
eka 2
eki 7
eks 3
ent 6
eri 5
hab 6
hek 5
him 4
hir 3
hit 3
hun 2
hus 6
iek 3
iik 6
ike 11
iki 2
ikt 6
ima 4
imt 6
ina 3
ini 11
inu 7
iri 16
isk 2
itb 6
iti 5
itm 3
itt 5
ius 7
kan 4
ken 4
kit 9
kst 6
kti 5
mab 4
nar 3
nen 2
nik 11
nun 2
nur 2
nus 5
rie 3
rik 8
rir 6
ris 2
riu 7
rur 4
sek 2
ska 4
ske 4
sks 3
skt 2
sku 2
tba 4
tbi 2
tit 5
tti 5
unk 2
unu 2
uri 2
urk 4
urr 5
use 2
usk 13
usm 4
ust 3
3-cgram-Tokens Final
aan 2
ban 8
bin 2
eat 4
euk 3
iek 4
ike 8
ira 6
kan 2
ken 11
kta 3
kur 2
mta 6
nke 2
nre 4
nta 6
rba 3
rke 4
rra 5
smi 4
sta 3
ste 6
tin 5
tir 5
tmi 3
3-cgram-Tokens Total
aan 2
aba 4
abi 12
ane 2
anr 4
arb 3
ari 3
aru 4
baa 2
ban 10
bar 4
bim 6
bin 13
bir 6
bit 2
eat 4
eka 2
eki 7
eks 3
ent 6
eri 5
euk 3
hab 6
hek 5
him 4
hir 3
hit 3
hun 2
hus 6
iek 7
iik 6
ike 19
iki 2
ikt 6
ima 4
imt 6
ina 3
ini 11
inu 7
ira 6
iri 16
isk 2
itb 6
iti 5
itm 3
itt 5
ius 7
kan 6
kea 4
ken 15
ker 2
keu 3
kie 4
kii 6
kit 9
kst 6
kta 3
kti 5
kur 2
lab 6
lek 2
lin 4
lir 3
lur 5
mab 4
mta 6
nar 6
nen 2
ner 3
nik 11
nke 2
nre 4
nta 6
nun 2
nur 2
nus 9
rba 3
rie 3
rik 8
rir 6
ris 2
riu 7
rke 4
rra 5
rur 4
sek 2
sha 6
she 5
shi 10
shu 8
sin 6
sir 4
ska 4
ske 4
sks 3
skt 2
sku 2
smi 4
sta 3
ste 6
tba 4
tbi 2
tik 2
tin 5
tir 5
tit 5
tmi 3
tti 5
unk 2
unu 2
uri 2
urk 4
urr 5
use 2
usk 13
usm 4
ust 3
Witten-Bell maasang nas invaeleydi gokaeyndang kna inela vi kraogeres a'vi na andnaki ys Jelinek-Mercer maasang wog gorrerred weogdang slkeemi ys anderbaelidora weogd. Ardeyn anderbaelidora weogd kaeyels vi slere ineyn na iniraomidora a'ne Good-Touring dydomide:
shusekan linurike 2 1
sinarba narikta 2 2
siriusmi sinuskste 2 3
tikitban shunusktin 2 4
Eloki Jelinek-Mercer maasang, ddeyr anvaelvi slog urger nas elawer urger n-gram mageel. Ne didi ys n'ynrd maasang mesas areyn ne Kneser-Ney maasang. Ard areyn na sdreora ys invaeleydi gokaeyndang weene ne kamvanidora ys slog urger nas elawer urger n-gram mageeleyn areyn veyoeld an in naveel mnaner, sid ar, eyang in mirganiel godroveydora andeis in miomeym elokeeloaas godroveydora:
2 AAN
12 ABA
3 ABE
18 ABI
2 ABT
3 ABU
3 AKA
2 AKE
2 ALA
5 ALI
3 ALU
2 AMI
2 ANA
5 ANB
4 ANE
2 ANI
4 ANK
6 ANL
2 ANM
2 ANN
5 ANR
11 ANS
5 ANT
4 ANU
5 ARB
9 ARI
2 ARM
2 ART
6 ARU
12 ASH
3 ASI
4 ATB
9 ATI
2 ATN
2 ATS
3 ATU
2 BAA
2 BAB
2 BAE
2 BAI
2 BAL
26 BAN
5 BAR
5 BAS
2 BBI
3 BEK
3 BEN
3 BER
3 BIK
7 BIM
23 BIN
6 BIR
3 BIT
2 BUR
3 BUS
5 EAT
3 EBA
2 EBE
4 EBI
2 EIR
5 EKA
3 EKB
2 EKE
16 EKI
2 EKN
4 EKS
6 EKU
3 EMI
3 ENA
4 ENB
3 ENE
4 ENI
2 ENK
2 ENR
7 ENS
9 ENT
4 ENU
4 ERA
7 ERI
5 ERU
11 ESH
3 EUK
6 HAB
5 HAT
2 HEI
6 HEK
4 HIM
11 HIR
3 HIT
5 HUN
3 HUR
14 HUS
2 IAN
2 IAT
2 IBB
8 IEK
2 IEN
6 IIK
2 IIR
3 IIT
2 IKA
3 IKB
27 IKE
9 IKI
7 IKT
12 IKU
6 IMA
3 IMB
2 IMI
2 IMK
2 IMR
8 IMT
6 INA
3 INB
12 INI
6 INK
4 INL
3 INR
7 INS
2 INT
8 INU
6 IRA
2 IRD
37 IRI
2 IRK
3 IRR
3 IRS
2 IRT
2 ISE
2 ISH
8 ISI
4 ISK
3 ISS
3 ISU
13 ITA
7 ITB
15 ITI
4 ITK
3 ITM
3 ITS
6 ITT
2 ITU
7 IUS
5 KAB
9 KAN
3 KAR
2 KAT
8 KBA
4 KBI
4 KEA
5 KEB
4 KEK
2 KEL
3 KEM
19 KEN
6 KER
10 KES
3 KEU
4 KIE
6 KII
5 KIK
2 KIN
2 KIR
2 KIS
13 KIT
5 KKE
2 KNU
4 KSH
7 KST
4 KTA
9 KTI
3 KUL
3 KUN
14 KUR
8 KUS
7 LAB
2 LEK
5 LIN
3 LIR
3 LIT
6 LUR
5 MAB
2 MAN
3 MBA
2 MEN
2 MIB
4 MIK
2 MIM
7 MIS
5 MIT
2 MRA
7 MTA
2 MUS
3 NAN
10 NAR
7 NBA
2 NBE
5 NBI
4 NEK
4 NEN
3 NER
18 NIK
2 NIM
2 NIR
4 NIS
3 NIT
7 NKE
4 NKI
2 NKU
3 NLA
6 NLI
2 NLU
3 NNI
6 NRE
4 NRI
21 NSH
8 NTA
5 NTE
3 NTI
3 NUN
4 NUR
17 NUS
7 RAB
3 RAK
5 RAL
2 RAN
2 RAR
2 RAS
2 RAT
9 RBA
2 RBI
7 REK
2 REM
3 REN
2 RIA
6 RIE
7 RII
11 RIK
10 RIR
7 RIS
14 RIT
8 RIU
6 RKE
2 RKI
3 RMI
7 RRA
2 RRE
4 RSH
2 RSK
3 RTI
5 RUN
9 RUR
2 RUS
2 SAN
3 SEK
2 SER
12 SHA
10 SHE
20 SHI
22 SHU
3 SIK
7 SIN
5 SIR
3 SIS
8 SKA
4 SKB
6 SKE
3 SKI
3 SKK
2 SKR
9 SKS
5 SKT
4 SKU
5 SMI
3 STA
10 STE
2 SUN
8 SUS
5 TAB
4 TAK
4 TAL
4 TAM
3 TAN
5 TAR
8 TAS
5 TAT
4 TBA
6 TBI
4 TEB
2 TEI
3 TEK
4 TEN
2 TER
3 TES
9 TIK
4 TIM
15 TIN
13 TIR
6 TIT
4 TKU
4 TMI
3 TSH
2 TSK
6 TTI
4 TUS
2 UKB
2 ULA
3 UNA
3 UNB
5 UNI
2 UNK
5 UNU
5 URB
9 URI
5 URK
2 URM
3 URN
6 URR
3 URS
3 URU
4 USA
4 USE
6 USH
4 USI
33 USK
5 USM
3 UST
5 USU
Ddene inne sleveriel user maasang inelgarosmeyn sid inne nad wogeelyr eyeg. An greeriel, Kneser-Ney maasang areyn negirges ineyn ne vdyd an maasang an brikdoke. Rar InHR, vedder maasang inelgarosmeyn miyr armbravdy ne wars errar ridi (WER) eyb a'1% invaeleydi:
baan: 2
baen: 1
baesk: 1
baib: 1
bair: 1
banenta: 2
banusu: 1
barurke: 4
barurra: 1
batuskba: 1
bauk: 1
bekirita: 1
bekunbin: 1
benekta: 1
bentin: 1
berabu: 1
beritkur: 1
beruniku: 1
binantin: 1
biniken: 11
bitikban: 1
bititban: 2
buranu: 1
buririsk: 1
busatbin: 1
dian: 1
doth: 1
kabanba: 1
kabantir: 1
kabikike: 1
kabiku: 1
kanu: 1
karisis: 1
katuskba: 1
keat: 4
kenariku: 1
kenisis: 1
keriskur: 2
kerusktin: 1
keuk: 3
kiat: 1
kiek: 4
kiike: 6
kinaniku: 1
kiriste: 1
kunatbin: 1
kuskitkur: 1
labirira: 6
labusan: 1
larurban: 1
lekitbin: 2
likusan: 1
linunbin: 1
linunke: 2
linurike: 2
liriekste: 3
lisuskste: 1
litaban: 1
litimban: 1
liturtir: 1
lurissk: 1
lurra: 5
lusitkur: 1
mabiman: 1
marimra: 1
mekuske: 1
menabiku: 1
menursk: 1
merunan: 1
mikarta: 1
mikikiku: 1
mimiritir: 1
misikke: 1
mitabta: 1
mitaniku: 1
mitarre: 1
mitenita: 1
munisike: 1
musabtin: 1
musiritin: 1
narikta: 3
nekurban: 1
neriktin: 3
nikatiku: 1
nikimra: 1
nikitbin: 1
nimankur: 1
nirianste: 1
nirienke: 1
nisisu: 1
nuseran: 1
nusimba: 1
nuskanre: 4
nuskurita: 1
rabekban: 1
rabenban: 1
rarinen: 1
ratikban: 1
rekisen: 1
renarke: 1
ririimta: 1
ririitra: 1
riserba: 1
ritarbin: 1
ritarmi: 1
ritatmi: 1
ritikike: 1
ritiriis: 1
ritusktir: 1
runike: 1
rurikba: 1
rurunba: 1
sartin: 1
shabimta: 6
shanensk: 1
shatimban: 1
shatimke: 1
shatiritin: 1
shatitsk: 1
shatusste: 1
sheat: 1
sheir: 1
sheit: 1
shekittir: 5
shekusktin: 1
shenekban: 1
shikitiku: 1
shimaban: 4
shinarban: 1
shiriatke: 1
shiriekita: 1
shirienis: 1
shiriirire: 1
shiriirita: 1
shiriitiku: 1
shiriitsk: 1
shiriurmi: 1
shiriusta: 3
shititmi: 3
shunarsk: 1
shunimtir: 1
shunusbin: 1
shunusktin: 2
shuritike: 1
shururbin: 1
shurusmi: 1
shusekan: 2
shuskabra: 1
shuskanba: 1
shuskanta: 1
shuskarmi: 1
shuskenta: 4
shuskinkur: 1
shuskuskre: 1
shususkke: 1
shususksk: 1
sinarba: 3
sinittir: 1
sinuskste: 3
siriusmi: 4
sunitkur: 1
sunuren: 1
susekkur: 1
susiken: 1
sususke: 1
tarekis: 1
tatimkur: 1
tatinke: 1
teab: 1
teib: 1
teir: 1
tekiktin: 1
tekuriku: 1
tenekbin: 1
tenuskke: 1
terunre: 1
tikitban: 2
Maasang aeknoreydy ga nad inggneeyn arseydy sleyk ineyn aabok ur daseli momidk sid sliireeyn vedwere n'arianang nas n'evieleyidora gidi. Kibdeyrang aabok neelides elrag rnagi gebregrekody wieyn rord andrageykes vyr Kneser nas Steinbiss wog ardeyn aabok modeyne mageel. Ne mageel kradianeyn mnayr wars vies n-grameyn weene eik ys ddem wieyn arianes eyang in gorrerred karbeyeyn wog in gorrerred aabok nas kamvanes an in elaneir wias:
e barurke biniken shususkke lurra biniken shatu
kke lurra biniken shatusste sunuren tenekbin b
ske kiike biniken terunre kiek mikarta kanu sh
ban sheat biniken kabantir shenekban . nerikt
linurike biniken shunusbin shatimke ratikban
larurban biniken shatitsk siriusmi sinuskste b
skba baib biniken riserba sinittir tikitban sh
tbin keuk biniken runike rurikba shekittir rab
skur baen biniken shuskanba shuritike siriusmi
tiku teib biniken buranu shururbin labirira kat
buririsk biniken bitikban shuskinkur . narik
Ddes Iyer edinel andrageykes ineydamidok keleyderang a'belod in elirgi karbeyeyn anda aabok bekorok karbari. Ddeyr eyes in slombeli kamvanidora ys anveri gakeymred rnereyrekody andeis ys eyang ne Elidred Slemnadok Angeang (ElHO) mesag, zok areyn weelel knaws an anrarmidora nedroeviel neeirk. Bellegarda edinel andrageykes ElHO a'ElM ineyn Elidred Slemnadok Naielasoeyn (ElHI):
"banenta",2,7,948.5
"barur",5,5,793.5
"barurke",4,7,876.8
"biniken",5,7,607.4
"bititban",2,8,587.5
"kaban",2,5,246
"kabik",2,5,523.5
"keriskur",2,8,105.5
"kiike",6,5,286.4
"labirira",6,8,553.3
"lekitbin",2,8,33
"linun",3,5,503.5
"linunke",2,7,39
"linurike",2,8,212.5
"liriekste",3,9,500.4
"lurra",5,5,660.1
"narikta",3,7,476.6
"neriktin",2,8,70.5
"nuskanre",4,8,348.9
"ririi",2,5,274.5
"ritar",2,5,181
"shabimta",6,8,753.1
"shati",4,5,209.1
"shatim",2,6,226.5
"shekittir",5,9,388.5
"shimaban",3,8,104.1
"shiri",11,5,631.3
"shirie",2,6,326.5
"shirii",4,6,522.4
"shiriiri",2,8,373.5
"shiriit",2,7,570
"shiriu",4,6,729.4
"shiriusta",3,9,815.9
"shititmi",3,8,850.2
"shunus",3,6,419.9
"shunusktin",2,10,315.5
"shuru",2,5,789.5
"shusekan",2,8,212.5
"shusk",10,5,613.3
"shuska",4,6,643.3
"shuskan",2,7,641.5
"shuskenta",4,9,646.1
"shususk",2,7,732
"sinarba",3,7,506.7
"sinuskste",3,9,805.1
"siriusmi",4,8,560.9
"tikitban",2,8,315.5
