EMNLP 2011 でひときわ目を引くタイトル "Cooooooooooooooollllllllllllll!!!!!!!!!!!!!! Using Word Lengthening to Detect Sentiment in Microblogs" (Brody & Diakopoulos) は、twitter 上での表現を評判分析に使う話。
Cooooooooooooooollllllllllllll のような文字を必要以上に繰り返す twitter でよくみる表現に注目して、このような表現が意見や評価を述べるときによく使われるという点を主張しているのがポイント。
まあ直感的にもうなずける分析だと思う。
論文の中では、そのためにまず Cooooooooooooooollllllllllllll -> cool のように正規化を行っている。
提案されている手法がとてもシンプルでおもしろそうだったので、その正規化だけちょっと試してみたよ、というのがこの記事。
論文に書かれている正規化の手順は次の通り。
- 同じ文字の繰り返しを全て 1 文字につぶした圧縮形を作る(niiiice -> nice, reallllly -> realy)
- 同じ圧縮形ごとにグルーピング
- 3文字以上の文字の繰り返しがないグループは除外
- 各グループの中で、もっとも高頻度の語をそのグループの標準形とする
うん、これなら手元にデータさえあればさくっと試せる。
なぜか、twitter StreamAPI から収集した、(ほぼ全件)英語なツイート 188万件(URL, mention, ハッシュタグ除外済み)が手元にあるので、そいつでさっくり試してみよう。
自然言語処理には Ruby が一番、というお告げがあったので、Ruby で書いてみたのがこちら。スクリプト名はもちろん cooool.rb .
#!/usr/bin/env ruby # -*- coding: utf-8 -*- def downcase(st) st.tr("A-Z\u00c0-\u00d6\u00d8-\u00de\u01e0", "a-z\u00e0-\u00f6\u00f8-\u00fe\u01e1") end map = Hash.new{|h,k| h[k] = Hash.new(0)} open(ARGV[0], "rb:utf-8") do |f| while line = f.gets line.scan(/[A-Za-z\u00c0-\u024f\u1e00-\u1eff]{2,}/) do |word| word = downcase(word) condensed = word.chars.inject("") do |w, c| if w[-1] != c then w + c else w end end map[condensed][word] += 1 end end end map.to_a.sort.each do |condensed, group| list = group.to_a.select{|x| x[1] >= 10} next if list.length <= 1 next unless list.any?{|x| x[0] =~ /(.)\1{2}/} canonical = list.max_by{|x| x[1]} ammount = list.inject(0){|s,x| s+x[1]} puts "#{canonical[0]}(#{canonical[1]}/#{ammount}) : [#{list.map{|x|x[0]}.sort.join(',')}]" end
論文には処理の簡略化やノイズ低減のため、「非アルファベットは除外」「1文字単語は除外」「10回以上出現した単語のみ」「小文字化」などなど書かれていたので基本その方針で。
アルファベットって、あれだよね、一番狭い意味でラテン文字全般ってことだよね、という多言語脳で Basic Latin だけではなく、Laten-1 Supplement と Latin Extended A/B/Additional まで入ってたりする。Extended C と D は多言語脳でも使われているのまだ見たこと無いので、さすがに入れてない。
と大きな口を叩く割りには、小文字化が Laten-1 Supplement までしか対応していないところが情けない(苦笑)。
そのうちラテン文字全般用の小(大)文字化作りたいと思っているんだけど、Extended B がカオスなんだよなあ。
で、189万件のツイートに対して実行すると、392個の正規化グループが得られた。抜粋して掲載しようかと思ったが途中でめんどくさくなったので、末尾にべたっと貼っておく。
おもしろいのは aye や hmm などの感嘆語や lol, lmao, lmfao のようなネットスラングに近い表現(ワロスwwwwww とかいう時に使う acronyms)もちゃんとと抽出できている。
tommorow,tommorrow,tomorow,tomorrow,tomorrowwww,tomorrrow のどれが正解でしょう? とか言われたら人間様だとちょっと迷うところだけど、ちゃんと正しい tomorrow が標準形として……んー、なんかじっと見てると、tomorrow で正しいかどうか自信なくなってきたw
ちなみに手順の3番目の「3文字以上の文字の繰り返しがないグループは除外」は単なる typo の類を除外するため、だそうだ。
英語は高々2文字までの繰り返ししかないからこれでうまくいくのかな?
試しにはずしてみると、mary と marry とかが同じグループに入ったりして、おーなるほどと腑に落ちた。
でも soon のグループに son が入ってたりするんだけどね(苦笑)。こういうのは辞書を持つしかないのかなあ、せっかくここまでシンプルなルールだけで来ているのにそれだとつまらないなあ&英語だけだとそれでもいいけど、多くの言語について適用することを考えると嬉しくないなあ、とつらつら考えていたが、「標準形より短い表現はグループから除外」というシンプルなルールを追加するだけでおおむね行けるんじゃあないかという気がしてきた。要検証だけど。
で、せっかくラテン文字全般を食えるように書いたんだし、英語にこだわる必要も全くないんだから、他の言語についてもやりたいよねえ、と思ってたら、なぜかたまたま手元に(ほぼ)ポルトガル語なツイート 203万件とか(ほぼ)スペイン語とか(ほぼ)オランダ語とかもあったりするので、こいつらももりっと食わせてみたら、やっぱり全然傾向が違っておもしろかったりするんだけど、そっちの結果はわざわざ貼るのはやめておく。
aaa(328/624) : [aa,aaa,aaaa] again(15918/15957) : [again,againn,againnn] agh(59/87) : [agh,aghh,aghhh] ah(2529/6104) : [aaaah,aaaahhhh,aaah,aaahh,aaahhh,aaahhhh,aah,aahh,aahhh,ah,ahh,ahhh,ahhhh,ahhhhh,ahhhhhh,ahhhhhhh,ahhhhhhhh,ahhhhhhhhh,ahhhhhhhhhh,ahhhhhhhhhhh,ahhhhhhhhhhhh,ahhhhhhhhhhhhh] aha(587/706) : [aha,ahaa,ahaaa,ahha] all(93634/95743) : [al,all,alll,allll,alllll,allllll,alllllll,allllllll,alllllllll,allllllllll] alone(3767/3791) : [alone,alonee,aloneee] already(12072/12123) : [already,alreadyy,alreadyyy] always(22064/22118) : [allways,always,alwayss,alwaysss] am(31607/31640) : [am,amm,ammm] amazing(7366/7429) : [amazing,amazingg,amazinggg,amazingggg,amazinggggg] and(280218/280633) : [aaaaaaand,aaaaaand,aaaaand,aaaand,aaand,aand,and,andd,anddd,andddd,anddddd,annd,annnd,annnnd,annnnnd,annnnnnd] are(90489/90539) : [are,aree,areee] as(50111/74362) : [as,ass,asss,assss,asssss,assssss,asssssss] asleep(2360/2370) : [asleeep,asleep] aww(2903/8099) : [aaw,aaww,aawww,aw,aww,awww,awwww,awwwww,awwwwww,awwwwwww,awwwwwwww,awwwwwwwww,awwwwwwwwww,awwwwwwwwwww] away(9775/9801) : [away,awayy,awayyy] awe(423/582) : [awe,awee,aweee,aweeee,awwe,awwwe] awesome(9419/9460) : [awesome,awesomee,awesomeee,awesomeeee] awh(281/532) : [awh,awhh,awhhh,awhhhh,awwh,awwhh,awwwh,awwwwh] ay(467/717) : [ay,ayy,ayyy,ayyyy,ayyyyy,ayyyyyyy] aye(1697/2657) : [aye,ayee,ayeee,ayeeee,ayeeeee,ayeeeeee,ayeeeeeee,ayye,ayyee,ayyeee,ayyeeee,ayyye,ayyyee,ayyyeee,ayyyeeee,ayyyye,ayyyyeee,ayyyyeeee,ayyyyye,ayyyyyye] ayo(229/261) : [ayo,ayoo,ayooo] bb(851/877) : [bb,bbb] baby(11789/12023) : [babby,baby,babyy,babyyy,babyyyy,babyyyyy] back(46040/46281) : [baaaack,baack,back,backk,backkk,backkkk] bad(16765/17180) : [baaaad,bad,badd,baddd,badddd,baddddd,baddddddd] be(133742/134252) : [be,bee,beee,beeee] bed(9037/9079) : [bed,bedd,beddd] been(31629/32486) : [beeeen,beeen,been,ben] best(22557/22606) : [best,bestt,besttt,besttttt] better(21898/21929) : [beter,better,bettter] bih(236/317) : [bih,bihh,bihhh,bihhhh] birthday(9620/9755) : [birthday,birthdayy,birthdayyy,birthdayyyy,birthdayyyyy,birthdayyyyyy] bitch(15306/15525) : [biiitch,bitch,bitchh,bitchhh,bitchhhh,bitchhhhh,bitchhhhhh] bitches(6163/6281) : [bitches,bitchess,bitchesss,bitchessss,bitchesssss] blah(959/1114) : [blah,blahh,blahhh,blahhhh,blahhhhh] boo(3744/4383) : [bo,boo,booo,boooo,booooo,boooooo,booooooo,boooooooo,booooooooo] boi(670/771) : [boi,boii,boiii,boiiii] bored(4177/4285) : [bored,boredd,boreddd,boredddd,boreddddd] boy(7739/7934) : [boooy,boy,boyy,boyyy,boyyyy,boyyyyy] bro(6595/6703) : [bro,broo,brooo,broooo] bruh(2026/2097) : [bruh,bruhh,bruhhh] burr(54/94) : [bur,burr,burrr] but(101214/102924) : [but,butt,buttt,butttt,buttttt,buut,buuut,buuuut,buuuuut] bye(1634/1738) : [bye,byee,byeee,byeeee,byeeeeee] cc(1461/1477) : [cc,ccc] call(15215/15432) : [cal,call,calll] chill(3158/3332) : [chil,chill,chilll,chillll,chilllll,chillllll,chilllllll,chillllllll] chillin(2106/2139) : [chilin,chillin,chillinnn,chilllin] climax(35/46) : [climax,climaxxx] cool(11519/11721) : [col,coll,cool,coool,cooool,coooool] come(25723/25781) : [come,comee,comeee,comme] crazy(10084/10348) : [craaaazy,craaazy,craazy,crazy,crazyy,crazyyy,crazyyyy,crazyyyyy,crazzy] ctfu(2468/2913) : [ctfu,ctfuu,ctfuuu,ctfuuuu,ctfuuuuu,ctfuuuuuu] ctfup(25/69) : [ctfup,ctfupp,ctfuppp,ctfuup] cute(6751/6926) : [cute,cutee,cuteee,cuteeee,cuteeeee,cuuute,cuuuuute] cuz(9633/9754) : [cuz,cuzz,cuzzz] dd(335/462) : [dd,ddd,dddd,ddddd,dddddd] dam(766/933) : [dam,damm,dammm] damn(17798/18384) : [daaaaamn,daaaamn,daaamn,daamn,dammn,damn,damnn,damnnn,damnnnn,damnnnnn,damnnnnnn,damnnnnnnn] dang(1355/1418) : [dang,dangg,danggg] dawg(680/743) : [dawg,dawgg,dawggg,dawgggg] day(47893/48108) : [day,dayy,dayyy,dayyyy,dayyyyy] days(12462/12530) : [days,dayss,daysss,dayssss] dead(5003/5017) : [dead,deaddd] def(3026/3423) : [def,deff,defff] die(4018/4036) : [die,dieee] do(74612/75105) : [do,doo,dooo,doooo,dooooo,doooooo] doe(2382/2512) : [doe,doee,doeee,doeeee] done(12153/12210) : [done,donee,doneee] down(23214/23310) : [down,downn,downnn,downnnn,downnnnn] dude(7187/7232) : [dude,dudee,dudeee,duuude] duh(681/899) : [duh,duhh,duhhh,duhhhh,duhhhhh] dying(1661/1684) : [dying,dyingg,dyinggg] ee(92/164) : [ee,eee,eeee,eeeee,eeeeee] early(7797/7839) : [early,earlyy,earlyyy,earlyyyy] eff(562/694) : [ef,eff,efff] eh(875/1283) : [eh,ehh,ehhh,ehhhh,ehhhhh] eek(71/132) : [eeeeek,eeeek,eeek,eek,ek] em(7055/7177) : [eem,em,emm,emmm] ep(674/713) : [eeep,eep,ep] er(632/838) : [er,err,errr,errrr] errbody(146/222) : [erbody,errbody,errrbody] ever(18055/18224) : [ever,everr,everrr,everrrr,everrrrr,everrrrrr] everyone(13999/14021) : [everyone,everyonee,everyoneee] ew(769/1894) : [eew,eewww,ew,eww,ewww,ewwww,ewwwww,ewwwwww,ewwwwwww,ewwwwwwww,ewwwwwwwww] exactly(3241/3252) : [exactly,exactlyyy] excited(5718/5746) : [excited,excitedd,exciteddd] ff(381/391) : [ff,fff] face(8796/8835) : [face,facee,faceee] facebook(8059/8071) : [facebook,faceboook] far(6148/6168) : [far,farr,farrr] feel(21634/23726) : [feeel,feel,fel,fell] finally(6998/7023) : [finallly,finally,finaly] fine(4026/4092) : [fine,finee,fineee,fineeee] fire(3378/3392) : [fire,fireee] food(8803/8861) : [food,foood,fooood,foooood] fool(2242/2309) : [fol,foll,fool,foool] follow(18521/18621) : [folllow,follow,followw,followww,folow] followers(7730/7779) : [followers,followerss,followersss,folowers] forever(3649/3689) : [forever,foreverrr,foreverrrr,foreverrrrrrr] free(14899/15010) : [fre,free,freee,freeee,freeeee,freeeeeee] friday(7750/7786) : [friday,fridayy,fridayyy,fridayyyy] fuck(27265/27768) : [fuck,fuckk,fuckkk,fuckkkk,fuckkkkk,fuckkkkkk,fuckkkkkkk,fuuck,fuuuck,fuuuuck,fuuuuuck,fuuuuuuck] fuk(542/680) : [fuk,fukk,fukkk] fun(11326/11465) : [fun,funn,funnn,funnnn] funny(10058/10122) : [funnny,funny,funnyy,funnyyy] gah(222/351) : [gah,gahh,gahhh,gahhhh] game(15379/15415) : [game,gamee,gameee] gee(244/416) : [ge,gee,geee] geesh(117/129) : [geeesh,geesh] getting(17100/17136) : [geting,getting,gettting] girl(19888/20208) : [giiirl,girl,girll,girlll,girllll,girlllll,girllllll,girlllllll,girllllllll,girrl,girrrl,girrrrl,girrrrrl] girls(9973/10007) : [girls,girlss,girlsss] go(57260/58024) : [go,goo,gooo,goooo,gooooo,goooooo,gooooooo,goooooooo,gooooooooo] good(60485/76173) : [god,godd,goddd,godddd,goddddd,good,goodd,goood,gooood,goooood,gooooood,goooooood,gooooooood] goodmorning(901/941) : [goodmorning,goodmorningg,goodmorninggg,gooodmorning] goodnight(1906/1928) : [goodnight,goodnightt,gooodnight] gonna(21802/22130) : [gona,gonna,gonnna,goona] gosh(740/772) : [gosh,goshh,goshhh] gotta(14957/15161) : [gota,gotta,gottta] gr(481/1099) : [gr,grr,grrr,grrrr,grrrrr,grrrrrr,grrrrrrr,grrrrrrrr,grrrrrrrrr] great(23865/23895) : [great,greatt,greattt] guess(10114/10188) : [gues,guess,guesss,guessss] guh(157/213) : [guh,guhh,guhhh] gurl(507/531) : [gurl,gurll,gurlll] guys(10853/10908) : [guys,guyss,guysss,guyssss] hh(71/135) : [hh,hhh] ha(5453/5877) : [ha,haa,haaa,haaaa,haaaaa,haaaaaaa,hha] haha(17292/17751) : [haaha,haha,hahaa,hahaaa,hahaaaa,hahha,hahhaa,hhaha] hahaha(4784/5026) : [hahaaha,hahaha,hahahaa,hahahaaa,hahahha,hahahhaa,hahhaha,hhahaha] happy(17063/17180) : [happpppy,happppy,happpy,happy,happyy,happyyyy,happyyyyy] hard(11333/11363) : [hard,hardd,harddd] hate(19506/19575) : [hate,hatee,hateee,hateeee] have(107172/107244) : [haave,have,havee,haveee] hell(10847/11250) : [heel,hel,hell,helll,hellll,helllll,hellllll,helllllll,hellllllll] hello(2711/2839) : [heello,helllo,hello,helloo,hellooo,helloooo,hellooooo,helloooooo,helo] her(40731/40849) : [her,herr,herrr,herrrr,herrrrr] here(35081/35254) : [here,heree,hereee,hereeee,hereeeee,herre] hey(12325/13260) : [heeeeeey,heeeeey,heeeey,heeey,heey,heeyy,heeyyy,hey,heyy,heyyy,heyyyy,heyyyyy,heyyyyyy,heyyyyyyy,heyyyyyyyy] hi(3377/3470) : [hi,hii,hiii,hiiii,hiiiiii] him(22421/22486) : [him,himm,himmm,himmmm] hmm(1594/4211) : [hhmm,hhmmm,hm,hmm,hmmm,hmmmm,hmmmmm,hmmmmmm,hmmmmmmm,hmmmmmmmm] ho(660/981) : [ho,hoo,hooo,hoooo] home(19433/19570) : [home,homee,homeee,homeeee,homeeeee,homme] hot(8457/8782) : [hoot,hot,hott,hottt,hotttt,hottttt,hotttttt] house(13147/13197) : [house,housee,houseee] huge(2381/2392) : [huge,huuuuge] huh(2558/2614) : [huh,huhh,huhhh,huhhhh] hum(180/240) : [hum,humm,hummm] hungry(3239/3287) : [hungry,hungryy,hungryyy,hungryyyy] hurr(60/100) : [hur,hurr,hurrr] ii(1370/1669) : [ii,iii,iiii,iiiii] idk(6035/6093) : [idk,idkk,idkkk] ie(166/176) : [ie,ieee] iight(648/1122) : [ight,ightt,iight,iiight] ill(6057/6374) : [il,ill,illl] im(40458/40615) : [iim,im,imm,immm,immmm] in(258182/258594) : [iin,in,inn,innn,innnn,innnnn,innnnnn] is(248954/249189) : [iis,is,iss,isss,issss,isssss] ish(1689/1778) : [ish,ishh,ishhh,issh] it(261451/261853) : [iit,it,itt,ittt,itttt,ittttt,itttttt,ittttttt] its(41454/41532) : [its,itss,itsss] jeez(174/184) : [jeeez,jeez] jus(6495/7189) : [jus,juss,jusss] kk(184/228) : [kk,kkk] keep(14494/14513) : [keeep,keep] kill(4165/4210) : [kil,kill,killl,killll] kno(5438/5500) : [kno,knoo,knooo] know(61663/62142) : [knooooow,knoooow,knooow,knoow,know,knoww,knowww,knowwww,knowwwww,knowwwwww,knowwwwwww] ll(32811/32844) : [ll,lll] late(6133/6301) : [late,latee,lateee,latte] later(5202/5327) : [later,laterr,laterrr,latter] lawd(472/517) : [lawd,lawdd,lawddd,lawwwd] lego(156/405) : [legggooo,leggo,leggoo,leggooo,leggoooo,lego,legoo,legooo,legoooo] lies(1430/1444) : [lies,liesss] life(26108/26220) : [life,lifee,lifeee,lifeeee,lifeeeee] like(103103/103301) : [liike,like,likee,likeee,likeeee,llike] live(14970/15007) : [live,livee,liveeeee] lmao(26905/33244) : [lmaaao,lmaaaooo,lmaao,lmaaoo,lmaaooo,lmaaoooo,lmao,lmaoo,lmaooo,lmaoooo,lmaooooo,lmaoooooo,lmaooooooo,lmaoooooooo,lmaooooooooo,lmaoooooooooo,lmaooooooooooo,lmaoooooooooooo,lmaooooooooooooo] lmaort(51/66) : [lmaooort,lmaort] lmbo(830/951) : [lmbo,lmboo,lmbooo,lmboooo,lmbooooo] lmfao(7302/11047) : [lmfaaaooo,lmfaao,lmfaaoo,lmfao,lmfaoo,lmfaooo,lmfaoooo,lmfaooooo,lmfaoooooo,lmfaooooooo,lmfaoooooooo,lmfaooooooooo,lmfaoooooooooo,lmfaooooooooooo,lmfaoooooooooooo,lmfaooooooooooooo,lmmfao,lmmfaooo,lmmfaoooo,lmmfaooooo] look(20814/20874) : [lok,look,loook] lol(141118/143131) : [llol,lol,loll,lolll,lollll,lolllll,lollllll,lolllllll,lool,loool,looool,loooool,looooool,loooooool,looooooool,looooooooool] long(13624/13870) : [long,longg,longgg,longggg,longgggg,looong,loooong,looooong,loooooong,looooooong] love(66003/67048) : [looooooove,loooooove,looooove,loooove,looove,loove,love,lovee,loveee,loveeee,loveeeee,loveeeeee,loveeeeeee,loveeeeeeee] loves(2992/3004) : [loves,lovesss] lls(7202/7623) : [llls,lls,llss,llsss,llssss,llsssss,ls] lucky(2359/2392) : [lucky,luckyy,luckyyy,luckyyyy] mm(684/2103) : [mm,mmm,mmmm,mmmmm,mmmmmm,mmmmmmm,mmmmmmmm,mmmmmmmmm] mad(7879/8245) : [maaad,mad,madd,maddd,madddd,maddddd,madddddd] man(25158/26345) : [maaaaaan,maaaaan,maaaan,maaan,maaannn,maan,maannn,man,mann,mannn,mannnn,mannnnn,mannnnnn,mannnnnnn,mannnnnnnn,mannnnnnnnn] mane(725/766) : [mane,manee,maneee] matters(727/740) : [matters,mattters] me(174872/176244) : [me,mee,meee,meeee,meeeee,meeeeee,meeeeeee,meeeeeeee,meeeeeeeee,meeeeeeeeee] mean(11371/11409) : [mean,meann,meannn,meannnn] meh(477/534) : [meeh,meh,mehh,mehhh] mhm(142/439) : [mhm,mhmm,mhmmm,mhmmmm,mmhm,mmhmm,mmhmmm,mmmhmm,mmmhmmm] mind(8041/8064) : [mind,mindd,minddd] miss(13518/13739) : [mis,miss,misss,missss,misssss] missed(4141/4152) : [missed,misssed] money(12263/12331) : [money,moneyy,moneyyy,moneyyyy,mooney] more(39710/40122) : [moore,more,moree,moreee,moreeee,moreeeee] morning(16953/17042) : [morning,morningg,morninggg,morningggg] move(6076/6086) : [move,moveee] movie(8703/8738) : [movie,moviee,movieee] muah(269/310) : [muah,muahh,muahhh,muahhhh] much(25808/25875) : [much,muchh,muchhh] my(268093/268418) : [mmy,my,myy,myyy,myyyy,myyyyy] na(1028/1222) : [na,naa,naaa,naaaa] nah(3098/3804) : [naaaah,naaah,naah,naahh,nah,nahh,nahhh,nahhhh,nahhhhh,nahhhhhh] nasty(1758/1782) : [nasty,nastyyy,nastyyyy] naw(1979/2166) : [naw,naww,nawww,nawwww,nawwwwww] need(42395/42574) : [ned,need,neeed,neeeed,neeeeed,neeeeeed] never(28470/28532) : [never,neverr,neverrr,neverrrr] nice(12517/12852) : [nice,nicee,niceee,niceeee,niceeeee,niceeeeee,niice,niiice,niiiice,niiiiice] nigga(14929/15039) : [niga,nigga,niggaa,niggaaa,niggaaaa,niggga] niggas(10009/10036) : [niggas,niggass,nigggas] night(25826/25916) : [night,nightt,nighttt,nightttt] no(65017/66841) : [no,noo,nooo,noooo,nooooo,noooooo,nooooooo,noooooooo,nooooooooo,noooooooooo,nooooooooooo,noooooooooooo] nope(2083/2127) : [nope,nopee,nopeee] not(96102/96199) : [not,nott,nottt,notttt,nottttt] nothing(11432/11461) : [nothing,nothingg,nothinggg] now(71107/71427) : [nooow,now,noww,nowww,nowwww,nowwwww,nowwwwww] ooo(452/1265) : [oo,ooo,oooo,ooooo,oooooo,ooooooo,oooooooo,ooooooooo,oooooooooooo] odee(103/190) : [ode,odee,odeee] of(230107/262833) : [of,off,offf,offff,offfff,offffff,offfffff,oof] oh(26801/31716) : [oh,ohh,ohhh,ohhhh,ohhhhh,ohhhhhh,ohhhhhhh,ohhhhhhhh,ohhhhhhhhh,ohhhhhhhhhh,ooh,oohh,oohhh,oohhhh,oooh,ooohh,ooohhh,ooohhhh,ooohhhhh,ooooh,oooohh,oooohhh,oooohhhh,oooooh,ooooooh,oooooooh] ok(12879/13058) : [ok,okk,okkk,okkkk,ook,oook] okay(6674/6989) : [okaaay,okaay,okay,okayy,okayyy,okayyyy,okayyyyy,okayyyyyy,okkay] omfg(705/728) : [omfg,omfgg,omfggg] omg(9311/9756) : [omg,omgg,omggg,omgggg,omggggg,omgggggg,omggggggg] omgee(31/43) : [omgee,omgeee] on(206597/206795) : [on,onn,onnn,onnnn,onnnnn,onnnnnn,oon] one(58954/59057) : [one,onee,oneee,oneeee] op(364/567) : [ooop,oop,op,opp] oops(644/1233) : [oooops,ooops,oops,opps,ops] or(48182/48244) : [or,orr,orrr,orrrr] ou(218/244) : [ou,ouu,ouuu] out(89493/89698) : [out,outt,outtt,outttt,outtttt] over(25467/25548) : [over,overr,overrr,overrrr,overrrrr] ow(279/574) : [ow,oww,owww,owwww,owwwww,owwwwww,owwwwwww] pp(90/104) : [pp,ppp] partay(20/30) : [partay,partayyy] party(11105/11187) : [party,partyy,partyyy,partyyyy] people(42011/42052) : [people,peoplee,peopleee] pf(73/101) : [pf,pff,pfff] pfft(92/163) : [pfffft,pffft,pfft,pft] phone(14348/14437) : [phone,phonee,phoneee] pissed(2022/2032) : [pissed,pisssed] please(16001/16332) : [please,pleasee,pleaseee,pleaseeee,pleaseeeee,pleaseeeeee,pleaseeeeeeee,pleeease,pleeeease] plz(1414/1527) : [plz,plzz,plzzz,plzzzz,pplz] pretty(10722/10746) : [prettty,pretty,prettyy] psh(119/273) : [psh,pshh,pshhh,pshhhh,pssh,psssh] pst(98/215) : [psssst,pssst,psst,pst] purr(26/53) : [pur,purr,purrr] qqq(10/20) : [qq,qqq] rr(83/97) : [rr,rrr] ready(11657/11718) : [ready,readyy,readyyy] real(17386/17490) : [real,reall,realll,reallll] really(35539/35841) : [realllllly,reallllly,realllly,reallly,realllyyy,really,reallyy,reallyyy,reallyyyy,realy,reeeeeally] right(38413/38661) : [right,rightt,righttt,rightttt,righttttt,riight,riiight,riiiight,riiiiight] room(7301/7404) : [rom,room,rooom] ss(266/280) : [ss,sss] same(13606/13648) : [same,samee,sameee] say(27496/27567) : [say,sayy,sayyy,sayyyy] school(17766/17814) : [school,schoool,schooool] see(43069/43780) : [se,see,seee,seeee,seeeee,sse] sexy(4008/4104) : [sexxxy,sexxy,sexy,sexyy,sexyyy,sexyyyy] sh(1001/1446) : [sh,shh,shhh,shhhh,shhhhh,shhhhhh,shhhhhhh,ssh] sheeit(12/24) : [sheeeit,sheeit] sheesh(707/828) : [sheeeeesh,sheeeesh,sheeesh,sheesh,sheeshh] shi(77/159) : [shi,shii,shiii] shid(188/403) : [shid,shidd,shiddd,shidddd,shiid,shiiid] shit(41202/41950) : [shiiiiiit,shiiiiit,shiiiit,shiiit,shiit,shiittt,shit,shitt,shittt,shitttt,shittttt,shitttttt,shittttttt] sho(376/422) : [sho,shoo,shooo] shore(2458/2495) : [shore,shoree,shoreee,shoreeeee] shot(3020/5213) : [shooooot,shooot,shoot,shot] show(17148/17175) : [show,showw,showww] shower(3288/3312) : [shower,showerr,showerrr] sick(4093/4143) : [sick,sickk,sickkk,sickkkk] sigh(1607/1644) : [sigh,sighh,sighhh] sleep(13397/13533) : [sleeeeeep,sleeeeep,sleeeep,sleeep,sleep,sleepp] sleeping(2260/2270) : [sleeeping,sleeping] slow(2672/2710) : [slow,sloww,slowww,slowwww] smh(19935/20208) : [smh,smhh,smhhh,smhhhh,smhhhhh] so(108337/117684) : [so,soo,sooo,soooo,sooooo,soooooo,sooooooo,soooooooo,sooooooooo,soooooooooo,sooooooooooo,soooooooooooo,sooooooooooooo,soooooooooooooo,sso] soon(9197/13752) : [son,sonn,sonnn,sonnnn,soon,sooon,soooon,sooooon] song(11222/11280) : [song,songg,songgg,songggg] sorry(8854/8928) : [sorrrry,sorrry,sorry,sorryy,sorryyy] still(32446/32623) : [stil,still,stilll,stillll,stilllll] stop(15291/15433) : [stoop,stop,stopp,stoppp,stopppp] stoopid(83/95) : [stooopid,stoopid] stuff(7090/7112) : [stuf,stuff,stufff] suck(3239/3249) : [suck,suckkk] sucks(3452/3475) : [sucks,suckss,sucksss] summer(7300/7312) : [summer,summmer] sure(13787/13829) : [sure,suree,sureee] swag(1814/2208) : [swag,swagg,swaggg,swagggg] swear(4480/4504) : [swear,swearr,swearrr] sweet(4619/4681) : [sweeeeet,sweeeet,sweeet,sweet] tt(821/831) : [tt,ttt] tell(17666/17779) : [tel,tell,telll] thank(13999/14032) : [thank,thankk,thankkk] thanks(22487/22743) : [thanks,thankss,thanksss,thankssss,thanksssss] thanx(793/834) : [thanx,thanxx,thanxxx] that(214810/215002) : [thaaat,thaat,that,thatt,thattt,thatttt,tthat] thats(10058/10100) : [thats,thatss,thatsss] the(610273/611271) : [the,thee,theee,theeee,theeeee,tthe] them(30712/30752) : [them,themm,themmm] then(28287/28343) : [then,thenn,thennn] there(45654/45709) : [there,theree,thereee,thereeee] thing(15995/16034) : [thing,thingg,thinggg] this(142148/142377) : [thiis,this,thiss,thisss,thissss,thisssss] tho(14382/14617) : [tho,thoo,thooo,thoooo,thooooo] though(12971/13038) : [though,thoughh,thoughhh,thoughhhh] time(50621/50788) : [time,timee,timeee,timeeee,timeeeee] tired(6322/6410) : [tired,tiredd,tireddd,tiredddd,tireddddd] to(520970/567306) : [to,too,tooo,toooo,tooooo,toooooo,tooooooo,toooooooo,tooooooooo,toooooooooo,tto] today(39687/39889) : [todaay,today,todayy,todayyy,todayyyy] tomorrow(16594/16974) : [tommorow,tommorrow,tomorow,tomorrow,tomorrowwww,tomorrrow] tonight(23201/23338) : [tonight,tonightt,tonighttt,tonightttt,tonighttttt] tru(440/486) : [tru,truu,truuu] true(8566/8733) : [true,truee,trueee,trueeee,trueeeee] tweet(13919/13941) : [tweeet,tweet] twitter(21558/21638) : [twiiter,twiter,twitter,twitterr,twittter] uu(287/361) : [uu,uuu,uuuu,uuuuu,uuuuuu] ugg(66/114) : [ug,ugg,uggg] ugh(4958/6627) : [uggggh,ugggh,uggghh,uggghhh,uggghhhh,uggh,ugghh,ugghhh,ugghhhh,ugh,ughh,ughhh,ughhhh,ughhhhh,ughhhhhh,ughhhhhhh,ughhhhhhhh,ughhhhhhhhhh,uugghh,uuugh] uh(1796/2749) : [uh,uhh,uhhh,uhhhh,uhhhhh,uhhhhhh] uhm(325/441) : [uhm,uhmm,uhmmm,uhmmmm] um(1787/4497) : [um,umm,ummm,ummmm,ummmmm,ummmmmm,ummmmmmm] up(95456/96012) : [up,upp,uppp,upppp,uppppp,upppppp,uppppppp] us(22283/22329) : [us,uss,usss] vvv(11/22) : [vv,vvv] very(11585/11623) : [verrrry,very,veryy,veryyy] www(4647/4760) : [ww,www] wack(1177/1197) : [wack,wackk,wackkkk] wah(270/527) : [waaaaah,waaaah,waaah,waah,wah,wahh,wahhh,wahhhh,wahhhhh,wahhhhhh] wait(16393/16465) : [waaaait,wait,waitt,waittt,waitttt] wanna(19391/20078) : [wana,wanna,wannaa,wannna] was(94640/94805) : [waas,was,wass,wasss] wassup(1357/1517) : [wasssup,wassup,wassupp,wassuppp,wasup] way(24664/25381) : [waaaaaay,waaaaay,waaaay,waaay,waaayyy,waay,way,wayy,wayyy,wayyyy,wayyyyy,wayyyyyy,wayyyyyyy] we(82120/82868) : [we,wee,weee,weeee,weeeee,wwe] week(15052/15070) : [weeek,week] well(25647/25883) : [wel,well,welll,wellll,welllll,wellllll,welllllll] welcome(4857/4906) : [welcome,welcomee,welcomeee] wha(249/358) : [wha,whaa,whaaa,whaaaa,whaaaaa] what(87949/88410) : [whaaaaaat,whaaaaat,whaaaat,whaaat,whaaattt,whaat,what,whatt,whattt,whatttt,whattttt,whatttttt] whatever(3142/3170) : [whatever,whateverrr,whateverrrr] where(23228/23251) : [where,wheree,whereee] whew(350/376) : [whew,wheww,whewww] who(46018/46303) : [who,whoo,whooo,whoooo,whooooo,whoooooo] whoa(1112/1189) : [whoa,whoaa,whoaaa,whoaaaa] whoop(556/583) : [whooop,whoop,whop] why(36194/36560) : [why,whyy,whyyy,whyyyy,whyyyyy,whyyyyyy,whyyyyyyy,whyyyyyyyy] wiff(131/261) : [wif,wiff,wifff] will(53135/53355) : [wil,will,willl,willll,willlll] woo(667/1247) : [wo,woo,wooo,woooo,wooooo,woooooo,wooooooo,woooooooo] woah(560/624) : [woah,woahh,woahhh,woahhhh,wooah] woohoo(350/442) : [wohoo,woohoo,woohooo,wooohoo,wooohooo,wooohoooo] woop(213/260) : [wooop,woop,wop] word(7466/7579) : [word,wordd,worddd,wordddd] work(27412/27476) : [work,workk,workkk,workkkk] world(13911/13948) : [world,worldd,worlddd,worldddd] woot(667/718) : [woooot,wooot,woot,wot] wow(8849/9278) : [wooooooow,woooooow,wooooow,woooow,wooow,woow,wow,woww,wowww,wowwww,wowwwww,wowwwwww] wtf(6464/6615) : [wtf,wtff,wtfff,wtffff,wtfffff,wtffffff] xx(414/660) : [xx,xxx,xxxx,xxxxx] xd(3052/3205) : [xd,xdd,xddd,xdddd] xi(53/85) : [xi,xii,xiii] ya(15598/15764) : [ya,yaa,yaaa,yaaaa,yaaaaa] yah(470/527) : [yah,yahh,yahhh,yahhhh] yall(8956/9370) : [yal,yall,yalll] yas(43/147) : [yas,yass,yasss,yassss,yasssss,yassssss] yay(3093/3894) : [yaaaaaay,yaaaaay,yaaaay,yaaay,yaay,yay,yayy,yayyy,yayyyy,yayyyyy,yayyyyyy,yayyyyyyy] ye(649/835) : [ye,yee,yeee,yeeee,yeeeee,yeeeeee] yea(9889/10638) : [yea,yeaa,yeaaa,yeaaaa,yeaaaaa,yeaaaaaa,yeeaa] yeah(16773/17960) : [yeaaaaah,yeaaaah,yeaaah,yeaaahh,yeaaahhh,yeaah,yeaahh,yeah,yeahh,yeahhh,yeahhhh,yeahhhhh,yeahhhhhh,yeahhhhhhh,yeeah,yeeeah,yeeeeah,yeeeeeah] yep(2225/2328) : [yep,yepp,yeppp] yes(15934/18670) : [yeeeees,yeesss,yes,yess,yesss,yessss,yesssss,yessssss,yesssssss,yessssssss,yesssssssss,yessssssssss,yesssssssssss,yessssssssssss,yesssssssssssss,yessssssssssssss] yessir(299/309) : [yessir,yesssir] yo(12988/14348) : [yo,yoo,yooo,yoooo,yooooo,yoooooo,yooooooo,yoooooooo,yooooooooo] you(366880/368322) : [yoooou,yooou,yoou,you,youu,youuu,youuuu,youuuuu,youuuuuu,youuuuuuu,youuuuuuuu,youuuuuuuuu] yu(6398/6525) : [yu,yuu,yuuu] yum(663/877) : [yum,yumm,yummm,yummmm,yummmmm,yummmmmm] yummy(695/707) : [yummmy,yummy] yup(2363/2679) : [yup,yupp,yuppp,yupppp,yuup,yuuup,yuuuup] zzz(55/278) : [zz,zzz,zzzz,zzzzz,zzzzzz,zzzzzzz,zzzzzzzz,zzzzzzzzz,zzzzzzzzzz,zzzzzzzzzzz]