{"id":712,"date":"2019-12-12T11:19:21","date_gmt":"2019-12-12T10:19:21","guid":{"rendered":"https:\/\/ispravi.me\/info\/?p=712"},"modified":"2023-01-04T12:21:16","modified_gmt":"2023-01-04T11:21:16","slug":"25-godina-haseka","status":"publish","type":"post","link":"https:\/\/ispravi.me\/info\/25-godina-haseka\/","title":{"rendered":"25 godina Ha\u0161eka"},"content":{"rendered":"<div class=\"clanak\">\n<p>AUTOR: <a href=\"https:\/\/www.fer.unizg.hr\/sandor.dembitz\">\u0160andor Dembitz<\/a><\/p>\n<p>OBJAVLJENO: <a href=\"https:\/\/hrcak.srce.hr\/jezik\">Jezik<\/a>, god. 66, br. 4-5, str. 138-150. Rad je primljen 2. travnja 2019., prihva\u0107en za tisak 7. listopada 2019. i nakon tiskanja pretvoren u ovaj oblik s dopu\u0161tenjem uredni\u0161tva Jezika.<\/p>\n<h3 class=\"naslov\">Uvod<\/h3>\n<p>Ime iz naslova \u010ditatelja vjerojatno najprije podsje\u0107a na Dobrog vojaka \u0160vejka a ponekog, mo\u017eda, i na Ljudevita Jonkea, prvog urednika Jezika, prevoditelja romana na hrvatski. Za razliku od \u010ceha Jaroslava Ha\u0161eka, koji je svoju svjetski poznatu satiru pisao tijekom i nakon Velikoga rata, hrvatski je vojnik \u0160vejk \u2013 pridjev \u201edobar\u201c namjerno je izostavljen \u2013 svoj Ha\u0161ek po\u010deo pisati tijekom Domovinskoga rata, te ga i dandanas dopisuje.<\/p>\n<p>Ha\u0161ek je pohrva\u0107eni oblik akronima <em>Hascheck<\/em>, izvedenog iz naziva Hrvatski akademski <em>spelling checker<\/em>, i ozna\u010dava jezgrenu komponentu mre\u017enog pravopisnog provjernika koji u razli\u010ditim oblicima, danas na adresi <a href=\"https:\/\/ispravi.me\/\">https:\/\/ispravi.me\/<\/a>, od 21. o\u017eujka 1994. stoji na raspolaganju svima koji \u017eele da im se tekst prije objavljivanja strojno provjeri.<\/p>\n<p>Danas, u guglzoiku, <em>spellchecking<\/em> nije posebno atraktivno podru\u010dje prirodnojezi\u010dnih tehnologija, \u0161to u doma\u0107im okvirima potvr\u0111uje spominjanje Ha\u0161eka u knjizi <a href=\"http:\/\/www.meta-net.eu\/whitepapers\/e-book\/croatian.pdf\">Hrvatski jezik u digitalnom dobu<\/a>, u kojoj mu je posve\u0107ena jedna jedina re\u010denica na 26. stranici: \u201e<em>On-line Hrvatski akademski spelling checker<\/em> (Hascheck) postoji od 1994. i jo\u0161 uvijek je u uporabi.\u201c U citiranoj se monografiji njezini autori, svi odreda barem jednom izabrani za \u010dlana-suradnika HAZU-a, iscrpno bave temama danas opredme\u0107enim u <em>Google Translateu<\/em> ili <em>Google Dictateu<\/em> itd. Jedino im je promakla \u010dinjenica da je Ha\u0161ek davna hrvatska anticipacija istih, ali \u0161to se tu mo\u017ee.<\/p>\n<p>\u010cemu uop\u0107e <em>on-line spellchecking<\/em>? U paleoguglzoiku, dok su se Amerikanci jo\u0161 intenzivno bavili pravopisnim provjernicima, o problemu je napisano i ovo:<\/p>\n<blockquote><p>\u201eRecept za izradu gula\u0161a od slona zapo\u010dinje s: prvo ulovi slona. Ako va\u0161 recept za izradu pravopisnog provjernika zapo\u010dinje s: prvo prona\u0111i sve valjane rije\u010di-razli\u010dnice u engleskom jeziku, vjerojatno \u0107ete brzo uvidjeti da je puno lak\u0161e napraviti ukusni gula\u0161 od slona.\u201c <a href=\"#_ftn1\" name=\"_ftnref1\">[1]<\/a><\/p><\/blockquote>\n<p>Lako je predo\u010div ameri\u010dki lovac, opremljen pu\u0161kom za uspavljivanje, kako si lovi svoga slona. \u0160to da radi njegov hrvatski parnjak, oboru\u017ean kamenom sjekirom, ako slu\u010dajno uspije o\u0161amutiti svoga mamuta? \u201eNa internet s njime, jer ina\u010de gula\u0161a nema!\u201c Da je ovo paleoliti\u010dko razmi\u0161ljanje bilo ispravno potvr\u0111uje \u010dinjenica da danas, osim Microsoftova pravopisnog provjernika za hrvatski, korisnicima hrvatskoga u stvarnosti za te svrhe jo\u0161 jedino Ha\u0161ek stoji na raspolaganju. Prije dvadesetak godina konvencionalnih <a href=\"https:\/\/ispravi.me\/info\/wp-content\/uploads\/2016\/07\/1997-03-WIN-INI.pdf\">hrvatskih pravopisnih provjernika<\/a> bilo je za na lopate bacati, ali nisu pre\u017eivjeli. Me\u0111unarodnim veletvrtkama \u0161aka jada ne mo\u017ee konkurirati po modelu: \u201evidjela \u017eaba kako potkivaju konja pa i sama digla nogu\u201c. Za takve izazove ipak treba malo soli u glavi. Da je izaziva\u010d strancima na koncu pokazao tko je tko na doma\u0107em bunji\u0161tu, potvr\u0111uje i <a href=\"https:\/\/ispravi.me\/info\/wp-content\/uploads\/2015\/12\/vidi-252-2017.pdf\">nedavna usporedba<\/a>.<\/p>\n<h3 class=\"naslov\">\u0160to je napravljeno?<\/h3>\n<p>Kako je Ha\u0161ek nastao, \u010demu sve slu\u017ei, kako radi i jo\u0161 puno toga zainteresirani \u010ditatelj Jezika mo\u017ee prona\u0107i u <a href=\"http:\/\/www.matica.hr\/kolo\/539\/strojna-obrada-hrvatskog-jezika-maarski-doprinosi-27748\/\">Kolu<\/a> i Filologiji <a href=\"#_ftn2\" name=\"_ftnref2\">[2]<\/a>. Stoga \u0107e ovdje ukratko biti prikazano samo ono \u0161to je u 25 godina napravljeno a da ima neku vrijednost.<\/p>\n<p>Ha\u0161ekov je rje\u010dnik od po\u010detnih 100.000 razli\u010dnica hrvatskog op\u0107ejezi\u010dnog fonda u 25 godina strogo nadziranog u\u010denja, nadziranoga radi o\u010duvanja preciznosti rje\u010dnika, narastao na:<\/p>\n<ul>\n<li>1.051.189 razli\u010dnica hrvatskog op\u0107ejezi\u010dnog fonda;<\/li>\n<li>957.620 razli\u010dnica hrvatskog posebnojezi\u010dnog, dominantno imenskog fonda;<\/li>\n<li>70.528 razli\u010dnica engleskog op\u0107ejezi\u010dnog fonda, u kojemu nema onih rije\u010di koje se identi\u010dno pi\u0161u u engleskome i hrvatskome, npr. atom ili zebra.<\/li>\n<\/ul>\n<p>Engleski leksik je uklju\u010den u Ha\u0161ekov rje\u010dnik jer je engleski jezik dana\u0161nja <em>lingua franca<\/em>. \u010cak se i u <a href=\"http:\/\/riznica.ihjj.hr\/index.hr.html\">Hrvatskoj jezi\u010dnoj riznici<\/a>, stomilijunskom dijakronijskom korpusu sa stoljetnim rasponom tekstova, koji su sastavili kroatisti, javlja 13.175 razli\u010dnica iz engleskog dijela Ha\u0161ekova rje\u010dnika (naju\u010destaliji je odre\u0111eni \u010dlan <em>the<\/em> s ukupno 7.988 pojavljivanja), koje tvore 0,4 % cjelovitoga korpusa Riznice. Uzimaju\u0107i u obzir i uko\u0161ene oblike engleskih rije\u010di tipa <em>rolla<\/em>, <em>rollu<\/em> itd., udio engle\u0161tine u Riznici penje se do 0,8 %, \u0161to odgovara razini zatipkovno-pravopisnih gre\u0161aka u njoj. Ina\u010de, Ha\u0161ekov bi rje\u010dnik, kada bi ga netko \u017eelio tiskati, tra\u017eio najmanje 3 standardna leksikografska sveska.<\/p>\n<p>U 25 godina usluzi je pristupljeno s 1.368.702 IP-adrese iz 177 vr\u0161nih internetskih domena, prete\u017eito zemalja. Prikaz opsega pru\u017eene usluge po vr\u0161nim domenama dan je u Dodatku ovom radu. Prema evidenciji HTTP kola\u010di\u0107a, tj. tragu koji svaki korisnik ostavlja za sobom nakon obavljene obrade, uslugu je koristilo oko milijun osoba. U Tablici 1. prikazana je ukupnost 25-godi\u0161njeg Ha\u0161ekovog uslu\u017eivanja najva\u017enijih vr\u0161nih domena s nekoliko bitnih parametara.<\/p>\n<table style=\"width: 592px; height: 367px;\" cellspacing=\"0\" cellpadding=\"0\">\n<colgroup>\n<col style=\"width: 150px;\" \/>\n<col style=\"width: 175px;\" span=\"4\" \/> <\/colgroup>\n<tbody>\n<tr>\n<td>Izvori\u0161ta prometa<\/td>\n<td>Obra\u0111eni korpus [pojavnica]<\/td>\n<td>Udio po izvori\u0161tima [%]<\/td>\n<td>Prosje\u010dno prekrivanje korpusa rje\u010dnikom [%]<\/td>\n<td>Prosje\u010dni udio zatipkovno-pravopisnih gre\u0161aka u korpusu [%]<\/td>\n<\/tr>\n<tr>\n<td>Hrvatska<\/td>\n<td><span class=\"broj\">6.313.123.913<\/span><\/td>\n<td><span class=\"postotak\">87,26<\/span><\/td>\n<td>98,47<\/td>\n<td>1,50<\/td>\n<\/tr>\n<tr>\n<td>BiH<\/td>\n<td><span class=\"broj\">460.404.455<\/span><\/td>\n<td><span class=\"postotak\">6,36<\/span><\/td>\n<td>97,17<\/td>\n<td>2,81<\/td>\n<\/tr>\n<tr>\n<td>Srbija <a href=\"#_ftn3\" name=\"_ftnref3\">[3]<\/a><\/td>\n<td><span class=\"broj\">58.941.003<\/span><\/td>\n<td><span class=\"postotak\">0,81<\/span><\/td>\n<td>97,31<\/td>\n<td>2,67<\/td>\n<\/tr>\n<tr>\n<td>Njema\u010dka<\/td>\n<td><span class=\"broj\">58.714.427<\/span><\/td>\n<td><span class=\"postotak\">0,81<\/span><\/td>\n<td>98,13<\/td>\n<td>1,83<\/td>\n<\/tr>\n<tr>\n<td>SAD<\/td>\n<td><span class=\"broj\">54.830.162<\/span><\/td>\n<td><span class=\"postotak\">0,76<\/span><\/td>\n<td>98,67<\/td>\n<td>1,31<\/td>\n<\/tr>\n<tr>\n<td>Ostala<\/td>\n<td><span class=\"broj\">289.082.052<\/span><\/td>\n<td><span class=\"postotak\">4,00<\/span><\/td>\n<td>97,68<\/td>\n<td>2,29<\/td>\n<\/tr>\n<tr>\n<td class=\"naglasi\">Ukupno<\/td>\n<td class=\"naglasi\"><span class=\"broj\">7.235.096.012<\/span><\/td>\n<td class=\"naglasi\"><span class=\"postotak\">100,00<\/span><\/td>\n<td class=\"naglasi\">98,34<\/td>\n<td class=\"naglasi\">1,62<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p class=\"opis\" style=\"text-align: center;\">Tablica 1.<\/p>\n<p>Obra\u0111eni korpus od 7,2 Gpojavnica (gigapojavnica) odgovara korpusu od 30 milijuna autorskih kartica teksta i 6 puta je ve\u0107i od \u201enajve\u0107eg hrvatskog korpusa hrWaC\u201c, kojim se na 35. stranici di\u010di <a href=\"http:\/\/www.meta-net.eu\/whitepapers\/e-book\/croatian.pdf\">uvodno citirana monografija<\/a>, \u0161to je samo jo\u0161 jedna potvrda da kod malih primjereno osmi\u0161ljeni pristupi znaju polu\u010diti bolje rezultate od nekriti\u010dkog slije\u0111enja velikih po \u017eabljem modelu.<\/p>\n<p>Ono \u0161to zabrinjava jest podatak koji upu\u0107uje da se hrvatski urednije pi\u0161e u SAD-u negoli u samoj Hrvatskoj (posljednji stupac Tablice 1.), ali to je pitanje kojim bi se morale pozabaviti hrvatske obrazovne vlasti. Poziv se opravdava \u010dinjenicom da su unatrag nekoliko posljednjih godina one bile vrlo izda\u0161ne u dodjeljivanju nagrade \u201eIvan Filipovi\u0107\u201c za zna\u010dajna ostvarenja u odgojno-obrazovnoj djelatnosti <a href=\"http:\/\/ihjj.hr\/stranica\/nagrade-i-priznanja\/25\/\">hrvatskim normativistima<\/a>, kojima je zada\u0107a hrvatske u\u010denike uputiti kako treba uredno pisati na hrvatskom jeziku. Nas sretnima \u010dine priznanja sljede\u0107e vrste:<\/p>\n<blockquote><p><em>Po\u0161tovani, pohvala za va\u0161u stranicu <\/em><a href=\"https:\/\/ispravi.me\/\"><em>https:\/\/ispravi.me\/<\/em><\/a><em>! Nisam izvorna govornica hrvatskog jezika i te\u0161ko mi pada pohvatati sve gramati\u010dke cake. Va\u0161a stranica mi daje samopouzdanja jer u\u010dim pri svakom pisanju. Hvala puno i samo naprijed! Lp, Tena<\/em> <a href=\"#_ftn4\" name=\"_ftnref4\">[4]<\/a><\/p><\/blockquote>\n<p>Ha\u0161ek je odavno prestao biti konvencionalni pravopisni provjernik. Ispravljanje gramati\u010dkih gre\u0161aka zapo\u010delo je mijenjanjem nepostoje\u0107eg glagolskog priloga pro\u0161log, primjerice \u201eslijediv\u0161i\u201c, u valjani glagolski prilog sada\u0161nji, tj. \u201eslijede\u0107i\u201c, i obrnuto, \u201eproslijede\u0107i\u201c u \u201eproslijediv\u0161i\u201c. \u010cak ni pismeni korisnici hrvatskoga nisu vi\u0161e sasvim sigurni, vjerojatno zbog gubitka aorista, odnosno imperfekta u svakodnevnoj uporabi, koji su hrvatski glagoli svr\u0161eni, a koji nesvr\u0161eni. Bavljenje \u201enekonvencionalnim gre\u0161kama\u201c nastavljeno je s kreiranjem hrvatskog n-gramskog sustava, koji je omogu\u0107io da se kontekstno prepoznaju, po potrebi i isprave, u\u010destale gramati\u010dke i stilske gre\u0161ke u pisanju na hrvatskome.<\/p>\n<p>Skupljanje i ure\u0111ivanje hrvatskih n-grama zapo\u010delo je, potaknuto <a href=\"https:\/\/en.wikipedia.org\/wiki\/Google_Translate\">projektom <em>Google Translate<\/em><\/a>, sredinom 2007. godine. N-gramski je sustav nu\u017ena podatkovna podloga za suo\u010davanje s izazovima kao \u0161to su strojno prevo\u0111enje, strojna pretvorba govora u tekst itd. U Tablici 2. nalazi se usporedni prikaz hrvatskoga s dva najve\u0107a Googleova n-gramska sustava s po\u010detka re\u010denoga projekta.<\/p>\n<table style=\"width: 585px; height: 300px;\" cellspacing=\"0\" cellpadding=\"0\">\n<colgroup>\n<col style=\"width: 180px;\" span=\"4\" \/> <\/colgroup>\n<tbody>\n<tr>\n<td style=\"border-left: none; border-top: none;\"><\/td>\n<td><a href=\"https:\/\/catalog.ldc.upenn.edu\/LDC2006T13\">Engleski<\/a><br \/>\nWaC<br \/>\n1,025 Tpojavnica<\/td>\n<td><a href=\"https:\/\/catalog.ldc.upenn.edu\/LDC2010T06\">Kineski<\/a><br \/>\nWaC<br \/>\n883 Gpojavnica<\/td>\n<td>Hrvatski<br \/>\nHa\u0161ekov korpus<br \/>\n7,2 Gpojavnica<\/td>\n<\/tr>\n<tr>\n<td>1-grami<\/td>\n<td><span class=\"broj\">13.588.391<\/span><\/td>\n<td><span class=\"broj\">1.616.150<\/span><\/td>\n<td><span class=\"broj\">5.757.442<\/span><\/td>\n<\/tr>\n<tr>\n<td>2-grami<\/td>\n<td><span class=\"broj\">314.843.401<\/span><\/td>\n<td><span class=\"broj\">281.107.315<\/span><\/td>\n<td><span class=\"broj\">265.171.603<\/span><\/td>\n<\/tr>\n<tr>\n<td>3-grami<\/td>\n<td><span class=\"broj\">977.069.902<\/span><\/td>\n<td><span class=\"broj\">1.024.642.142<\/span><\/td>\n<td><span class=\"broj\">918.083.221<\/span><\/td>\n<\/tr>\n<tr>\n<td>4-grami<\/td>\n<td><span class=\"broj\">1.313.818.354<\/span><\/td>\n<td><span class=\"broj\">1.348.990.533<\/span><\/td>\n<td><span class=\"broj\">1.390.001.665<\/span><\/td>\n<\/tr>\n<tr>\n<td>5-grami<\/td>\n<td><span class=\"broj\">1.176.470.663<\/span><\/td>\n<td><span class=\"broj\">1.256.043.325<\/span><\/td>\n<td><span class=\"broj\">1.463.796.046<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"naglasi\">Ukupno<\/td>\n<td class=\"naglasi\"><span class=\"broj\">3.795.790.711<\/span><\/td>\n<td class=\"naglasi\"><span class=\"broj\">3.912.399.465<\/span><\/td>\n<td class=\"naglasi\"><span class=\"broj\">4.042.809.977<\/span><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p class=\"opis\" style=\"text-align: center;\">Tablica 2.<\/p>\n<p>Google se poslu\u017eio cjelokupnim WWW-om kao tekstovnim repozitorijem, odnosno tzv. <em>Web as Corpus<\/em> (WaC) pristupom \u2013 isti je poslu\u017eio i za dobivanje maloprije spomenutoga \u201enajve\u0107eg hrvatskog korpusa\u201c \u2013 i \u010destotno\u0161\u0107u n-grama, primijeniv\u0161i tzv. <em>cut-off<\/em> kriterij, da bi dobio gore prikazane sustave. To u hrvatskom slu\u010daju ne mo\u017ee voditi do usporedivih rezultata, ali do usporedivih se rezultata dolazi ako se iskoriste Ha\u0161ekove obrade i leksi\u010dnost kao kriterij za uvr\u0161tavanje n-grama u bazu, tj. da su konstituenti svih n-grama rije\u010di s potvrdom u Ha\u0161ekovom rje\u010dniku. Valja napomenuti da preko 50 % unigrama u hrvatskom slu\u010daju tvore razli\u010dnice-brojevi, no ve\u0107 s n \u2265 2 udio n-grama s takvim konstituentima pada ispod 2 %.<\/p>\n<p>Ha\u0161ekov 25-godi\u0161nji dru\u0161tveni doprinos mo\u017ee se sa\u017eeti u sljede\u0107em:<\/p>\n<ol>\n<li>U\u0161te\u0111eno je oko 10.000 radnih godina sri\u010du\u0107ega \u010ditanja, koje bi se bez usluge potro\u0161ile radi otkrivanja i otklanjanja gre\u0161aka, neizostavnih pratiteljica nastajanja novoga teksta.<\/li>\n<li>Stvoren je hrvatski n-gramski sustav, podatkovna podloga nu\u017ena za uspje\u0161no suo\u010davanje s izazovima koji stoje pred hrvatskim jezi\u010dnim tehnolozima, \u010diji je opseg ve\u0107i od opsega svih knjiga koje su od Gutenberga do danas tiskane na hrvatskom jeziku.<\/li>\n<\/ol>\n<p>Kako je usluga <a href=\"https:\/\/ispravi.me\/\">https:\/\/ispravi.me\/<\/a> zapravo predlektoriranje, osmi\u0161ljena da bi se ure\u0111iva\u010du teksta olak\u0161ao i skratio najnekreativniji, a vrlo zamorni dio posla, izra\u010dun prvoga doprinosa polazi od:<\/p>\n<ul>\n<li>davna lektorska norma kretala se izme\u0111u 10 i 20 autorskih kartica teksta dnevno;<\/li>\n<li>radna godina prema europskom standardu broji 1.720 radnih sati, odnosno 215 radnih dana.<\/li>\n<\/ul>\n<p>Ha\u0161ek je obradi 30.000.000 autorskih kartica teksta, pa ra\u010dunajte.<\/p>\n<p>Opseg korpusa svi knjiga tiskanih od Gutenberga do 2010. godine broji 18,2 Tpojavnica <a href=\"#_ftn5\" name=\"_ftnref5\">[5]<\/a>, iz \u010dega slijedi procjene da sve knjige ikada tiskane na hrvatskome tvore korpus \u010diji opseg ne prema\u0161uje 20 Gpojavnica. Opseg hrvatskog n-gramskog sustava, mjeren pojavnicama, ra\u010duna se iz podataka posljednjega stupca Tablice 2. na sljede\u0107i na\u010din:<\/p>\n<p class=\"podnaslov\">\u2211<sup>5<\/sup><sub><em>i<\/em> = 1<\/sub> (<em>broj_i_grama<\/em>) \u00b7 (<em>i<\/em> + 1) = 20,2 Gpojavnica<\/p>\n<p>i na tome se temelji navedena veli\u010dina drugoga doprinosa.<\/p>\n<p>Ha\u0161ek je ovoliko opstao zahvaljuju\u0107i uplatama manje od jednog promila njegovih korisnika, koji ga rabe ili su ga rabili u profesionalne svrhe. Skrb o usluzi po\u010diva na le\u0111ima aktualnog dekana FER-a i njegovog umirovljenika, \u010dije je zdravlje dobrano naru\u0161eno. Sre\u0107om, obojica jo\u0161 di\u0161u.<\/p>\n<h3 class=\"naslov\">\u0160to nije napravljeno?<\/h3>\n<p>Vijest o postojanju hrvatskog n-gramskog sustava potaknula ja Francuze, koji rade na sustavu <a href=\"https:\/\/www.liglab.fr\/fr\/la-recherche\/plates-formes-du-lig\/plate-forme-pour-le-traitement-automatique-des-langues\">Ariane<\/a>, da predlo\u017ee da se njihov francusko-ruski par, razvijan od vremena kada je Francuska pod de Gaulleom napustila NATO, metodom samonadopunjavanja (engl. <em>bootstrapping<\/em>) pretvori u francusko-hrvatski par za strojno prevo\u0111enje. Prijedlog je djelovao zdravo, jer je nudio mogu\u0107nost da se u razumnom roku s malim ulaganjima do\u0111e do visokokvalitetnog sustava za strojno prevo\u0111enje s francuskoga na hrvatski, i obrnuto. O kakvoj se kvaliteti prevo\u0111enja razmi\u0161ljalo dovoljno govori podatak da je za <em>benchmarking<\/em>, tj. usporedbu pokazatelja kakvo\u0107e prevo\u0111enja, odabran Saint-Exup\u00e9ryjev <em>Le Petit Prince<\/em>, kod nas davno preveden od strane jedne Spli\u0107anke kao Mali princ, potom u izdanju iz 2011. preimenovan u Malog kraljevi\u0107a. Me\u0111utim, od zamisli se nije daleko stiglo, jer ni tra\u017eena sredstva za pokrivanje materijalnih tro\u0161kova projekta nisu odobrena. Za\u0161to?<\/p>\n<p>Hrvatska politika, bilo koje vrste, nikada nije ozbiljno shva\u0107ala <a href=\"https:\/\/www.itu.int\/newsarchive\/press\/PP98\/Documents\/Statement_Gore.html\">Digitalnu deklaraciju me\u0111uovisnosti<\/a><em>,<\/em> politi\u010dku najavu guglzoika napisanu od strane osobe koja je dobila Nobelovu nagradu za mir 2007. godine. Posebno je njezinu drugu to\u010dku:<\/p>\n<blockquote><p>\u201eMoramo prevladati na\u0161e jezi\u010dne barijere razvijaju\u0107i stvarnovremenske sustave za strojno govorno prevo\u0111enje, tako da svatko na svijetu mo\u017ee razgovarati s bilo kim drugim\u201c<\/p><\/blockquote>\n<p>ona do\u017eivljavala kao <em>science fiction<\/em>. Izravni dokazi s po\u010detka guglzoika za potkrjepu ove tvrdnje trebali bi se nalaziti u arhivima MZO-a, HAZU-a i IHJJ-a. Ne\u0161to svje\u017eiji, premda neizravni dokaz slijedi:<\/p>\n<ul>\n<li>iz adresnih raspona Hrvatskog sabora (IP \u2013adrese 194.152.219.0 &#8211; 194.152.219.255, odnosno 195.29.174.0 &#8211; 195.29.175.255) u 25 godina obra\u0111ena su 2.872 teksta koji su tvorili korpus od 864.479 pojavnica, od \u010dega je 99,94 % prometa ostvareno u posljednjih 15 mjeseci, od po\u010detka 2018. do konca o\u017eujka 2019.;<\/li>\n<li>iz adresnog raspona Europskog parlamenta (IP-adrese 136.173.0.0 &#8211; 136.173.255.255) Ha\u0161ek je od po\u010detka 2013. do konca o\u017eujka 2019. zaprimio na obradu 14.522 teksta koji su tvorili korpus od 2.122.054 pojavnice, s manje-vi\u0161e jednolikom razdiobom prometa u vremenu.<\/li>\n<\/ul>\n<p>Dostatno.<\/p>\n<p>U govornotehnolo\u0161kom segmentu (strojna tvorba govora, odnosno strojno pretvaranje govora u tekst) jednostavnija rje\u0161enja (strojna tvorba govora, upravljanje govorom) na hrvatskom tr\u017ei\u0161tu nude slovenske i srpske tvrtke, jer hrvatskih tvrtki, koje bi im konkurirale, jednostavno nema. No, pravo vrhnje u ovom podru\u010dju bere Newton Technologies Adria, lokalna podru\u017enica \u010de\u0161ke tvrtke, koja je nedavno <a href=\"https:\/\/pravosudje.gov.hr\/vijesti\/potpisan-ugovor-o-nabavi-programskog-rjesenja-za-pretvaranje-govora-u-tekst-s-pripadajucim-specijaliziranim-uredjajima\/19861\">Ministarstvu pravosu\u0111a RH<\/a> prodala sustav za pretvorbu kontinuiranoga govora u tekst \u201es pripadaju\u0107im specijaliziranim ure\u0111ajima za diktiranje za 800 korisnika\u201c za 33,5 milijuna kuna. Uzalud svi prijedlozi davno upu\u0107eni Hrvatskoj zakladi za znanost da je nastupilo vrijeme za pokretanje projekata ciljanih prema razvoju hrvatskih govornotehnolo\u0161kih proizvoda. Uzalud dokazivanja da se uporabljivi prototipovi sustava, kako za strojnu tvorbu govora <a href=\"#_ftn6\" name=\"_ftnref6\">[6]<\/a>, tako i za pretvaranje kontinuiranoga govora u tekst <a href=\"#_ftn7\" name=\"_ftnref7\">[7]<\/a>, dadu brzo napraviti, i to bez ikakvih financijskih ulaganja, samo temeljeno na dobrim doma\u0107im podatkovnim podlogama i radu ne doktoranada, ve\u0107 diplomanata. Izgleda da je u Hrvatskoj isplativije sufinancirati tu\u0111i nego poticati vlastiti tehnolo\u0161ki razvoj, \u010dak i kada je u pitanju jezik bez kojega bi Hrvatska bila tek zemljopisna odrednica. Valja napomenuti da su prije 25 godina \u010cesi i Hrvati dijelili istu razinu razvijenosti prirodnojezi\u010dnih tehnologija <a href=\"#_ftn8\" name=\"_ftnref8\">[8]<\/a>.<\/p>\n<h3 class=\"naslov\">Zaklju\u010dak<\/h3>\n<p>Prije 150 godina pokrenuta je izrada tzv. Akademijina rje\u010dnika, grandioznoga projekta koji je trajao preko 100 godina, da bi se pokazalo kako je hrvatski ravnopravan svim drugim europskim jezicima. U dana\u0161njoj su Europi svi jezici nazivno ravnopravni, no u stvarnosti su neki ne\u0161to ravnopravniji, kao u onoj poznatoj \u017eivotinjskoj farmi. Za male narode, njihovu kulturu i identitet, nu\u017eno je stoga da u 21. stolje\u0107u izbore, i putem jezi\u010dnih tehnologija, svoje mjesto pod suncem ravnopravnosti. Malo je podru\u010dja nad kojima danas mali narod mo\u017ee iskazivati potpuni suverenitet kao \u0161to je to njegov jezik.<\/p>\n<p>Jasno je da se od suvereniteta uvijek mo\u017ee odustajati, ako za to postoje valjani razlozi. Takva odustajanja imaju svoju cijenu i u pravilu po\u010divaju na politi\u010dkim procjenama. O cijenama je ovdje bilo ne\u0161to rije\u010di, a za politi\u010dke procjene Ha\u0161ekov autor nije mjerodavan. Mo\u017ee samo iskazati svoju bojazan da \u0107e se hrvatskom jeziku do konca 21. stolje\u0107a vratiti status <em>K\u00fcchensprachea<\/em><em>, <\/em>kakav je imao prije Akademijina rje\u010dnika, odustanu li Hrvati od razvoja jezi\u010dnih tehnologija za vlastiti jezik. Ovaj rad upu\u0107uje da je takav scenarij, na autorovu veliku \u017ealost, danas ve\u0107 na djelu. \u010cemu su se onda Strossmayer i toliki nakon njega uop\u0107e trudili, neki i ginuli?<\/p>\n<h3 class=\"naslov\">DODATAK<\/h3>\n<h4 class=\"podnaslov\">Prikaz opsega pru\u017eene usluge po vr\u0161nim domenama<\/h4>\n<p>Budu\u0107i da su nazivi vr\u0161nih domena uzeti iz ameri\u010dke baze, prikaz je pisan engleskim pravopisom.<\/p>\n<table class=\"ispis\" style=\"width: 587px; height: 5603px;\" cellspacing=\"0\" cellpadding=\"0\">\n<colgroup>\n<col style=\"width: 80px;\" \/>\n<col style=\"width: 260px;\" \/>\n<col style=\"width: 170px;\" span=\"3\" \/> <\/colgroup>\n<tbody>\n<tr>\n<td style=\"border-left: none; border-top: none;\"><\/td>\n<td>IP-domains (countries)<\/td>\n<td>#IP-addresses<\/td>\n<td>#Texts<\/td>\n<td>Corpus [tokens]<\/td>\n<\/tr>\n<tr>\n<td>1.<\/td>\n<td>Afghanistan<\/td>\n<td>14<\/td>\n<td>128<\/td>\n<td>10,907<\/td>\n<\/tr>\n<tr>\n<td>2.<\/td>\n<td>Albania<\/td>\n<td>665<\/td>\n<td>3,808<\/td>\n<td>652,605<\/td>\n<\/tr>\n<tr>\n<td>3.<\/td>\n<td>Algeria<\/td>\n<td>20<\/td>\n<td>40<\/td>\n<td>7,319<\/td>\n<\/tr>\n<tr>\n<td>4.<\/td>\n<td>Andorra<\/td>\n<td>6<\/td>\n<td>22<\/td>\n<td>5,172<\/td>\n<\/tr>\n<tr>\n<td>5.<\/td>\n<td>Angola<\/td>\n<td>2<\/td>\n<td>5<\/td>\n<td>194<\/td>\n<\/tr>\n<tr>\n<td>6.<\/td>\n<td>Anonymous Proxy<\/td>\n<td>20<\/td>\n<td>1,646<\/td>\n<td>330,606<\/td>\n<\/tr>\n<tr>\n<td>7.<\/td>\n<td>Argentina<\/td>\n<td>104<\/td>\n<td>492<\/td>\n<td>168,571<\/td>\n<\/tr>\n<tr>\n<td>8.<\/td>\n<td>Armenia<\/td>\n<td>7<\/td>\n<td>41<\/td>\n<td>13,557<\/td>\n<\/tr>\n<tr>\n<td>9.<\/td>\n<td>Asia\/Pacific Region<\/td>\n<td>11<\/td>\n<td>67<\/td>\n<td>13,578<\/td>\n<\/tr>\n<tr>\n<td>10.<\/td>\n<td>Australia<\/td>\n<td>738<\/td>\n<td>7,590<\/td>\n<td>1,869,227<\/td>\n<\/tr>\n<tr>\n<td>11.<\/td>\n<td>Austria<\/td>\n<td>7,019<\/td>\n<td>129,741<\/td>\n<td>25,148,812<\/td>\n<\/tr>\n<tr>\n<td>12.<\/td>\n<td>Azerbaijan<\/td>\n<td>13<\/td>\n<td>26<\/td>\n<td>2,868<\/td>\n<\/tr>\n<tr>\n<td>13.<\/td>\n<td>Bahrain<\/td>\n<td>4<\/td>\n<td>9<\/td>\n<td>279<\/td>\n<\/tr>\n<tr>\n<td>14.<\/td>\n<td>Bangladesh<\/td>\n<td>7<\/td>\n<td>18<\/td>\n<td>14,873<\/td>\n<\/tr>\n<tr>\n<td>15.<\/td>\n<td>Barbados<\/td>\n<td>5<\/td>\n<td>40<\/td>\n<td>2,865<\/td>\n<\/tr>\n<tr>\n<td>16.<\/td>\n<td>Belarus<\/td>\n<td>32<\/td>\n<td>78<\/td>\n<td>24,734<\/td>\n<\/tr>\n<tr>\n<td>17.<\/td>\n<td>Belgium<\/td>\n<td>1,608<\/td>\n<td>25,464<\/td>\n<td>5,409,281<\/td>\n<\/tr>\n<tr>\n<td>18.<\/td>\n<td>Belize<\/td>\n<td>7<\/td>\n<td>292<\/td>\n<td>41,935<\/td>\n<\/tr>\n<tr>\n<td>19.<\/td>\n<td>Bermuda<\/td>\n<td>1<\/td>\n<td>1<\/td>\n<td>41<\/td>\n<\/tr>\n<tr>\n<td>20.<\/td>\n<td>Bolivia<\/td>\n<td>10<\/td>\n<td>98<\/td>\n<td>47,783<\/td>\n<\/tr>\n<tr>\n<td>21.<\/td>\n<td>Bosnia and Herzegovina<\/td>\n<td>108,122<\/td>\n<td>1,491,045<\/td>\n<td>460,404,455<\/td>\n<\/tr>\n<tr>\n<td>22.<\/td>\n<td>Botswana<\/td>\n<td>1<\/td>\n<td>15<\/td>\n<td>10,887<\/td>\n<\/tr>\n<tr>\n<td>23.<\/td>\n<td>Bouvet Island<\/td>\n<td>1<\/td>\n<td>7<\/td>\n<td>42,037<\/td>\n<\/tr>\n<tr>\n<td>24.<\/td>\n<td>Brazil<\/td>\n<td>212<\/td>\n<td>975<\/td>\n<td>196,390<\/td>\n<\/tr>\n<tr>\n<td>25.<\/td>\n<td>British Virgin Islands<\/td>\n<td>3<\/td>\n<td>13<\/td>\n<td>2,784<\/td>\n<\/tr>\n<tr>\n<td>26.<\/td>\n<td>Brunei<\/td>\n<td>1<\/td>\n<td>1<\/td>\n<td>928<\/td>\n<\/tr>\n<tr>\n<td>27.<\/td>\n<td>Bulgaria<\/td>\n<td>306<\/td>\n<td>12,359<\/td>\n<td>1,272,561<\/td>\n<\/tr>\n<tr>\n<td>28.<\/td>\n<td>Burkina Faso<\/td>\n<td>1<\/td>\n<td>1<\/td>\n<td>19<\/td>\n<\/tr>\n<tr>\n<td>29.<\/td>\n<td>Burundi<\/td>\n<td>3<\/td>\n<td>16<\/td>\n<td>695<\/td>\n<\/tr>\n<tr>\n<td>30.<\/td>\n<td>Cambodia<\/td>\n<td>115<\/td>\n<td>695<\/td>\n<td>91,950<\/td>\n<\/tr>\n<tr>\n<td>31.<\/td>\n<td>Cameroon<\/td>\n<td>14<\/td>\n<td>22<\/td>\n<td>83,891<\/td>\n<\/tr>\n<tr>\n<td>32.<\/td>\n<td>Canada<\/td>\n<td>1,190<\/td>\n<td>43,247<\/td>\n<td>9,996,040<\/td>\n<\/tr>\n<tr>\n<td>33.<\/td>\n<td>Cape Verde<\/td>\n<td>2<\/td>\n<td>10<\/td>\n<td>63<\/td>\n<\/tr>\n<tr>\n<td>34.<\/td>\n<td>Chile<\/td>\n<td>58<\/td>\n<td>309<\/td>\n<td>124,996<\/td>\n<\/tr>\n<tr>\n<td>35.<\/td>\n<td>China<\/td>\n<td>371<\/td>\n<td>5,498<\/td>\n<td>1,344,131<\/td>\n<\/tr>\n<tr>\n<td>36.<\/td>\n<td>Colombia<\/td>\n<td>53<\/td>\n<td>428<\/td>\n<td>85,234<\/td>\n<\/tr>\n<tr>\n<td>37.<\/td>\n<td>Congo &#8211; Brazzaville<\/td>\n<td>1<\/td>\n<td>8<\/td>\n<td>717<\/td>\n<\/tr>\n<tr>\n<td>38.<\/td>\n<td>Congo &#8211; Kinshasa<\/td>\n<td>4<\/td>\n<td>30<\/td>\n<td>4,419<\/td>\n<\/tr>\n<tr>\n<td>39.<\/td>\n<td>Costa Rica<\/td>\n<td>22<\/td>\n<td>70<\/td>\n<td>12,677<\/td>\n<\/tr>\n<tr>\n<td>40.<\/td>\n<td>C\u00f4te d&#8217;Ivoire<\/td>\n<td>6<\/td>\n<td>78<\/td>\n<td>26,673<\/td>\n<\/tr>\n<tr>\n<td>41.<\/td>\n<td>Croatia<\/td>\n<td>1,155,346<\/td>\n<td>23,142,519<\/td>\n<td>6,313,123,913<\/td>\n<\/tr>\n<tr>\n<td>42.<\/td>\n<td>Cuba<\/td>\n<td>4<\/td>\n<td>4<\/td>\n<td>50<\/td>\n<\/tr>\n<tr>\n<td>43.<\/td>\n<td>Cura\u00e7ao<\/td>\n<td>1<\/td>\n<td>1<\/td>\n<td>125<\/td>\n<\/tr>\n<tr>\n<td>44.<\/td>\n<td>Cyprus<\/td>\n<td>47<\/td>\n<td>236<\/td>\n<td>51,506<\/td>\n<\/tr>\n<tr>\n<td>45.<\/td>\n<td>Czech Republic<\/td>\n<td>890<\/td>\n<td>35,282<\/td>\n<td>7,002,622<\/td>\n<\/tr>\n<tr>\n<td>46.<\/td>\n<td>Denmark<\/td>\n<td>564<\/td>\n<td>11,565<\/td>\n<td>1,799,119<\/td>\n<\/tr>\n<tr>\n<td>47.<\/td>\n<td>Dominican Republic<\/td>\n<td>4<\/td>\n<td>31<\/td>\n<td>1,114<\/td>\n<\/tr>\n<tr>\n<td>48.<\/td>\n<td>Ecuador<\/td>\n<td>17<\/td>\n<td>83<\/td>\n<td>10,697<\/td>\n<\/tr>\n<tr>\n<td>49.<\/td>\n<td>Egypt<\/td>\n<td>83<\/td>\n<td>652<\/td>\n<td>26,395<\/td>\n<\/tr>\n<tr>\n<td>50.<\/td>\n<td>El Salvador<\/td>\n<td>5<\/td>\n<td>99<\/td>\n<td>15,377<\/td>\n<\/tr>\n<tr>\n<td>51.<\/td>\n<td>Estonia<\/td>\n<td>1,503<\/td>\n<td>12,057<\/td>\n<td>3,123,082<\/td>\n<\/tr>\n<tr>\n<td>52.<\/td>\n<td>Ethiopia<\/td>\n<td>15<\/td>\n<td>116<\/td>\n<td>28,273<\/td>\n<\/tr>\n<tr>\n<td>53.<\/td>\n<td>Europe<\/td>\n<td>1,398<\/td>\n<td>96,952<\/td>\n<td>15,193,772<\/td>\n<\/tr>\n<tr>\n<td>54.<\/td>\n<td>Faroe Islands<\/td>\n<td>3<\/td>\n<td>11<\/td>\n<td>848<\/td>\n<\/tr>\n<tr>\n<td>55.<\/td>\n<td>Finland<\/td>\n<td>248<\/td>\n<td>4,546<\/td>\n<td>962,307<\/td>\n<\/tr>\n<tr>\n<td>56.<\/td>\n<td>France<\/td>\n<td>2,027<\/td>\n<td>109,255<\/td>\n<td>20,372,694<\/td>\n<\/tr>\n<tr>\n<td>57.<\/td>\n<td>French Polynesia<\/td>\n<td>4<\/td>\n<td>14<\/td>\n<td>6,946<\/td>\n<\/tr>\n<tr>\n<td>58.<\/td>\n<td>Gambia<\/td>\n<td>1<\/td>\n<td>1<\/td>\n<td>1<\/td>\n<\/tr>\n<tr>\n<td>59.<\/td>\n<td>Georgia<\/td>\n<td>35<\/td>\n<td>156<\/td>\n<td>27,077<\/td>\n<\/tr>\n<tr>\n<td>60.<\/td>\n<td>Germany<\/td>\n<td>17,675<\/td>\n<td>293,479<\/td>\n<td>58,714,427<\/td>\n<\/tr>\n<tr>\n<td>61.<\/td>\n<td>Ghana<\/td>\n<td>4<\/td>\n<td>5<\/td>\n<td>1,000<\/td>\n<\/tr>\n<tr>\n<td>62.<\/td>\n<td>Gibraltar<\/td>\n<td>1<\/td>\n<td>2<\/td>\n<td>444<\/td>\n<\/tr>\n<tr>\n<td>63.<\/td>\n<td>Greece<\/td>\n<td>357<\/td>\n<td>1,533<\/td>\n<td>477,706<\/td>\n<\/tr>\n<tr>\n<td>64.<\/td>\n<td>Grenada<\/td>\n<td>13<\/td>\n<td>40<\/td>\n<td>9,442<\/td>\n<\/tr>\n<tr>\n<td>65.<\/td>\n<td>Guadeloupe<\/td>\n<td>2<\/td>\n<td>2<\/td>\n<td>3,010<\/td>\n<\/tr>\n<tr>\n<td>66.<\/td>\n<td>Guatemala<\/td>\n<td>5<\/td>\n<td>49<\/td>\n<td>7,369<\/td>\n<\/tr>\n<tr>\n<td>67.<\/td>\n<td>Guernsey<\/td>\n<td>1<\/td>\n<td>3<\/td>\n<td>662<\/td>\n<\/tr>\n<tr>\n<td>68.<\/td>\n<td>Haiti<\/td>\n<td>1<\/td>\n<td>1<\/td>\n<td>163<\/td>\n<\/tr>\n<tr>\n<td>69.<\/td>\n<td>Honduras<\/td>\n<td>1<\/td>\n<td>1<\/td>\n<td>45<\/td>\n<\/tr>\n<tr>\n<td>70.<\/td>\n<td>Hong Kong SAR China<\/td>\n<td>175<\/td>\n<td>1,239<\/td>\n<td>215,751<\/td>\n<\/tr>\n<tr>\n<td>71.<\/td>\n<td>Hungary<\/td>\n<td>1,601<\/td>\n<td>18,159<\/td>\n<td>4,801,973<\/td>\n<\/tr>\n<tr>\n<td>72.<\/td>\n<td>Iceland<\/td>\n<td>62<\/td>\n<td>299<\/td>\n<td>118,901<\/td>\n<\/tr>\n<tr>\n<td>73.<\/td>\n<td>India<\/td>\n<td>329<\/td>\n<td>1,116<\/td>\n<td>334,232<\/td>\n<\/tr>\n<tr>\n<td>74.<\/td>\n<td>Indonesia<\/td>\n<td>158<\/td>\n<td>522<\/td>\n<td>157,330<\/td>\n<\/tr>\n<tr>\n<td>75.<\/td>\n<td>Iran<\/td>\n<td>30<\/td>\n<td>117<\/td>\n<td>21,279<\/td>\n<\/tr>\n<tr>\n<td>76.<\/td>\n<td>Iraq<\/td>\n<td>73<\/td>\n<td>151<\/td>\n<td>20,819<\/td>\n<\/tr>\n<tr>\n<td>77.<\/td>\n<td>Ireland<\/td>\n<td>2,098<\/td>\n<td>18,091<\/td>\n<td>4,936,897<\/td>\n<\/tr>\n<tr>\n<td>78.<\/td>\n<td>Isle of Man<\/td>\n<td>5<\/td>\n<td>59<\/td>\n<td>18,481<\/td>\n<\/tr>\n<tr>\n<td>79.<\/td>\n<td>Israel<\/td>\n<td>133<\/td>\n<td>430<\/td>\n<td>137,631<\/td>\n<\/tr>\n<tr>\n<td>80.<\/td>\n<td>Italy<\/td>\n<td>3,050<\/td>\n<td>49,308<\/td>\n<td>8,844,232<\/td>\n<\/tr>\n<tr>\n<td>81.<\/td>\n<td>Jamaica<\/td>\n<td>11<\/td>\n<td>37<\/td>\n<td>12,695<\/td>\n<\/tr>\n<tr>\n<td>82.<\/td>\n<td>Japan<\/td>\n<td>216<\/td>\n<td>1,792<\/td>\n<td>322,026<\/td>\n<\/tr>\n<tr>\n<td>83.<\/td>\n<td>Jersey<\/td>\n<td>1<\/td>\n<td>2<\/td>\n<td>190<\/td>\n<\/tr>\n<tr>\n<td>84.<\/td>\n<td>Jordan<\/td>\n<td>27<\/td>\n<td>66<\/td>\n<td>104,807<\/td>\n<\/tr>\n<tr>\n<td>85.<\/td>\n<td>Kazakhstan<\/td>\n<td>32<\/td>\n<td>167<\/td>\n<td>21,420<\/td>\n<\/tr>\n<tr>\n<td>86.<\/td>\n<td>Kenya<\/td>\n<td>34<\/td>\n<td>798<\/td>\n<td>101,094<\/td>\n<\/tr>\n<tr>\n<td>87.<\/td>\n<td>Kuwait<\/td>\n<td>37<\/td>\n<td>122<\/td>\n<td>55,197<\/td>\n<\/tr>\n<tr>\n<td>88.<\/td>\n<td>Kyrgyzstan<\/td>\n<td>6<\/td>\n<td>12<\/td>\n<td>5,744<\/td>\n<\/tr>\n<tr>\n<td>89.<\/td>\n<td>Laos<\/td>\n<td>19<\/td>\n<td>62<\/td>\n<td>12,999<\/td>\n<\/tr>\n<tr>\n<td>90.<\/td>\n<td>Latvia<\/td>\n<td>123<\/td>\n<td>1,118<\/td>\n<td>261,875<\/td>\n<\/tr>\n<tr>\n<td>91.<\/td>\n<td>Lebanon<\/td>\n<td>12<\/td>\n<td>34<\/td>\n<td>4,674<\/td>\n<\/tr>\n<tr>\n<td>92.<\/td>\n<td>Liberia<\/td>\n<td>1<\/td>\n<td>1,029<\/td>\n<td>284,667<\/td>\n<\/tr>\n<tr>\n<td>93.<\/td>\n<td>Libya<\/td>\n<td>5<\/td>\n<td>12<\/td>\n<td>4,655<\/td>\n<\/tr>\n<tr>\n<td>94.<\/td>\n<td>Liechtenstein<\/td>\n<td>12<\/td>\n<td>2,489<\/td>\n<td>366,166<\/td>\n<\/tr>\n<tr>\n<td>95.<\/td>\n<td>Lithuania<\/td>\n<td>2,236<\/td>\n<td>12,556<\/td>\n<td>2,950,112<\/td>\n<\/tr>\n<tr>\n<td>96.<\/td>\n<td>Luxembourg<\/td>\n<td>539<\/td>\n<td>4,412<\/td>\n<td>1,231,743<\/td>\n<\/tr>\n<tr>\n<td>97.<\/td>\n<td>Macau SAR China<\/td>\n<td>3<\/td>\n<td>8<\/td>\n<td>1,206<\/td>\n<\/tr>\n<tr>\n<td>98.<\/td>\n<td>Madagascar<\/td>\n<td>5<\/td>\n<td>8<\/td>\n<td>833<\/td>\n<\/tr>\n<tr>\n<td>99.<\/td>\n<td>Malawi<\/td>\n<td>14<\/td>\n<td>171<\/td>\n<td>692,180<\/td>\n<\/tr>\n<tr>\n<td>100.<\/td>\n<td>Malaysia<\/td>\n<td>98<\/td>\n<td>335<\/td>\n<td>68,028<\/td>\n<\/tr>\n<tr>\n<td>101.<\/td>\n<td>Maldives<\/td>\n<td>6<\/td>\n<td>8<\/td>\n<td>357<\/td>\n<\/tr>\n<tr>\n<td>102.<\/td>\n<td>Malta<\/td>\n<td>102<\/td>\n<td>924<\/td>\n<td>142,161<\/td>\n<\/tr>\n<tr>\n<td>103.<\/td>\n<td>Martinique<\/td>\n<td>1<\/td>\n<td>1<\/td>\n<td>1,310<\/td>\n<\/tr>\n<tr>\n<td>104.<\/td>\n<td>Mauritania<\/td>\n<td>2<\/td>\n<td>3<\/td>\n<td>2,790<\/td>\n<\/tr>\n<tr>\n<td>105.<\/td>\n<td>Mauritius<\/td>\n<td>20<\/td>\n<td>51<\/td>\n<td>6,205<\/td>\n<\/tr>\n<tr>\n<td>106.<\/td>\n<td>Mexico<\/td>\n<td>171<\/td>\n<td>1,320<\/td>\n<td>358,737<\/td>\n<\/tr>\n<tr>\n<td>107.<\/td>\n<td>Moldova<\/td>\n<td>106<\/td>\n<td>1,763<\/td>\n<td>499,313<\/td>\n<\/tr>\n<tr>\n<td>108.<\/td>\n<td>Monaco<\/td>\n<td>22<\/td>\n<td>390<\/td>\n<td>44,370<\/td>\n<\/tr>\n<tr>\n<td>109.<\/td>\n<td>Mongolia<\/td>\n<td>2<\/td>\n<td>9<\/td>\n<td>204<\/td>\n<\/tr>\n<tr>\n<td>110.<\/td>\n<td>Montenegro<\/td>\n<td>5,921<\/td>\n<td>74,412<\/td>\n<td>26,743,505<\/td>\n<\/tr>\n<tr>\n<td>111.<\/td>\n<td>Morocco<\/td>\n<td>59<\/td>\n<td>226<\/td>\n<td>42,278<\/td>\n<\/tr>\n<tr>\n<td>112.<\/td>\n<td>Mozambique<\/td>\n<td>4<\/td>\n<td>22<\/td>\n<td>6,768<\/td>\n<\/tr>\n<tr>\n<td>113.<\/td>\n<td>Myanmar (Burma)<\/td>\n<td>31<\/td>\n<td>625<\/td>\n<td>62,308<\/td>\n<\/tr>\n<tr>\n<td>114.<\/td>\n<td>Nepal<\/td>\n<td>21<\/td>\n<td>103<\/td>\n<td>26,308<\/td>\n<\/tr>\n<tr>\n<td>115.<\/td>\n<td>Netherlands<\/td>\n<td>2,299<\/td>\n<td>59,282<\/td>\n<td>15,222,549<\/td>\n<\/tr>\n<tr>\n<td>116.<\/td>\n<td>New Zealand<\/td>\n<td>104<\/td>\n<td>988<\/td>\n<td>188,779<\/td>\n<\/tr>\n<tr>\n<td>117.<\/td>\n<td>Nicaragua<\/td>\n<td>10<\/td>\n<td>18<\/td>\n<td>12,338<\/td>\n<\/tr>\n<tr>\n<td>118.<\/td>\n<td>Nigeria<\/td>\n<td>33<\/td>\n<td>2,015<\/td>\n<td>232,345<\/td>\n<\/tr>\n<tr>\n<td>119.<\/td>\n<td>North Macedonia<\/td>\n<td>1,653<\/td>\n<td>18,334<\/td>\n<td>4,433,953<\/td>\n<\/tr>\n<tr>\n<td>120.<\/td>\n<td>Norway<\/td>\n<td>360<\/td>\n<td>5,474<\/td>\n<td>1,982,203<\/td>\n<\/tr>\n<tr>\n<td>121.<\/td>\n<td>Oman<\/td>\n<td>115<\/td>\n<td>628<\/td>\n<td>48,591<\/td>\n<\/tr>\n<tr>\n<td>122.<\/td>\n<td>Pakistan<\/td>\n<td>17<\/td>\n<td>79<\/td>\n<td>4,449<\/td>\n<\/tr>\n<tr>\n<td>123.<\/td>\n<td>Palestinian Territories<\/td>\n<td>1<\/td>\n<td>1<\/td>\n<td>7<\/td>\n<\/tr>\n<tr>\n<td>124.<\/td>\n<td>Panama<\/td>\n<td>19<\/td>\n<td>231<\/td>\n<td>91,467<\/td>\n<\/tr>\n<tr>\n<td>125.<\/td>\n<td>Paraguay<\/td>\n<td>1<\/td>\n<td>1<\/td>\n<td>5<\/td>\n<\/tr>\n<tr>\n<td>126.<\/td>\n<td>Peru<\/td>\n<td>35<\/td>\n<td>224<\/td>\n<td>23,228<\/td>\n<\/tr>\n<tr>\n<td>127.<\/td>\n<td>Philippines<\/td>\n<td>95<\/td>\n<td>382<\/td>\n<td>51,338<\/td>\n<\/tr>\n<tr>\n<td>128.<\/td>\n<td>Pitcairn Islands<\/td>\n<td>1<\/td>\n<td>2<\/td>\n<td>249<\/td>\n<\/tr>\n<tr>\n<td>129.<\/td>\n<td>Poland<\/td>\n<td>2,358<\/td>\n<td>45,167<\/td>\n<td>12,304,620<\/td>\n<\/tr>\n<tr>\n<td>130.<\/td>\n<td>Portugal<\/td>\n<td>419<\/td>\n<td>3,151<\/td>\n<td>778,821<\/td>\n<\/tr>\n<tr>\n<td>131.<\/td>\n<td>Puerto Rico<\/td>\n<td>5<\/td>\n<td>40<\/td>\n<td>12,507<\/td>\n<\/tr>\n<tr>\n<td>132.<\/td>\n<td>Qatar<\/td>\n<td>93<\/td>\n<td>1,815<\/td>\n<td>494,898<\/td>\n<\/tr>\n<tr>\n<td>133.<\/td>\n<td>R\u00e9union<\/td>\n<td>2<\/td>\n<td>23<\/td>\n<td>1,959<\/td>\n<\/tr>\n<tr>\n<td>134.<\/td>\n<td>Romania<\/td>\n<td>567<\/td>\n<td>19,195<\/td>\n<td>3,749,730<\/td>\n<\/tr>\n<tr>\n<td>135.<\/td>\n<td>Russia<\/td>\n<td>512<\/td>\n<td>8,487<\/td>\n<td>1,759,307<\/td>\n<\/tr>\n<tr>\n<td>136.<\/td>\n<td>Rwanda<\/td>\n<td>2<\/td>\n<td>2<\/td>\n<td>90<\/td>\n<\/tr>\n<tr>\n<td>137.<\/td>\n<td>Saint Kitts and Nevis<\/td>\n<td>3<\/td>\n<td>40<\/td>\n<td>29,618<\/td>\n<\/tr>\n<tr>\n<td>138.<\/td>\n<td>Saint Lucia<\/td>\n<td>2<\/td>\n<td>2<\/td>\n<td>182<\/td>\n<\/tr>\n<tr>\n<td>139.<\/td>\n<td>Satellite Provider<\/td>\n<td>4<\/td>\n<td>11<\/td>\n<td>665<\/td>\n<\/tr>\n<tr>\n<td>140.<\/td>\n<td>Saudi Arabia<\/td>\n<td>53<\/td>\n<td>439<\/td>\n<td>67,714<\/td>\n<\/tr>\n<tr>\n<td>141.<\/td>\n<td>Senegal<\/td>\n<td>10<\/td>\n<td>38<\/td>\n<td>57,054<\/td>\n<\/tr>\n<tr>\n<td>142.<\/td>\n<td>Serbia<\/td>\n<td>9,676<\/td>\n<td>88,909<\/td>\n<td>58,941,003<\/td>\n<\/tr>\n<tr>\n<td>143.<\/td>\n<td>Seychelles<\/td>\n<td>62<\/td>\n<td>42,806<\/td>\n<td>6,526,067<\/td>\n<\/tr>\n<tr>\n<td>144.<\/td>\n<td>Sierra Leone<\/td>\n<td>1<\/td>\n<td>2<\/td>\n<td>6<\/td>\n<\/tr>\n<tr>\n<td>145.<\/td>\n<td>Singapore<\/td>\n<td>158<\/td>\n<td>956<\/td>\n<td>707,089<\/td>\n<\/tr>\n<tr>\n<td>146.<\/td>\n<td>Slovakia<\/td>\n<td>466<\/td>\n<td>9,813<\/td>\n<td>1,946,409<\/td>\n<\/tr>\n<tr>\n<td>147.<\/td>\n<td>Slovenia<\/td>\n<td>12,774<\/td>\n<td>246,846<\/td>\n<td>33,146,688<\/td>\n<\/tr>\n<tr>\n<td>148.<\/td>\n<td>South Africa<\/td>\n<td>78<\/td>\n<td>803<\/td>\n<td>225,069<\/td>\n<\/tr>\n<tr>\n<td>149.<\/td>\n<td>South Korea<\/td>\n<td>85<\/td>\n<td>323<\/td>\n<td>52,287<\/td>\n<\/tr>\n<tr>\n<td>150.<\/td>\n<td>South Sudan<\/td>\n<td>1<\/td>\n<td>6<\/td>\n<td>2,200<\/td>\n<\/tr>\n<tr>\n<td>151.<\/td>\n<td>Spain<\/td>\n<td>1,384<\/td>\n<td>13,014<\/td>\n<td>6,896,783<\/td>\n<\/tr>\n<tr>\n<td>152.<\/td>\n<td>Sri Lanka<\/td>\n<td>31<\/td>\n<td>46<\/td>\n<td>7,340<\/td>\n<\/tr>\n<tr>\n<td>153.<\/td>\n<td>Sudan<\/td>\n<td>6<\/td>\n<td>12<\/td>\n<td>1,249<\/td>\n<\/tr>\n<tr>\n<td>154.<\/td>\n<td>Suriname<\/td>\n<td>1<\/td>\n<td>1<\/td>\n<td>73<\/td>\n<\/tr>\n<tr>\n<td>155.<\/td>\n<td>Sweden<\/td>\n<td>1,829<\/td>\n<td>50,094<\/td>\n<td>7,935,319<\/td>\n<\/tr>\n<tr>\n<td>156.<\/td>\n<td>Switzerland<\/td>\n<td>1,647<\/td>\n<td>27,318<\/td>\n<td>8,642,473<\/td>\n<\/tr>\n<tr>\n<td>157.<\/td>\n<td>Syria<\/td>\n<td>5<\/td>\n<td>9<\/td>\n<td>302<\/td>\n<\/tr>\n<tr>\n<td>158.<\/td>\n<td>Taiwan<\/td>\n<td>56<\/td>\n<td>214<\/td>\n<td>64,589<\/td>\n<\/tr>\n<tr>\n<td>159.<\/td>\n<td>Tajikistan<\/td>\n<td>2<\/td>\n<td>2<\/td>\n<td>108<\/td>\n<\/tr>\n<tr>\n<td>160.<\/td>\n<td>Tanzania<\/td>\n<td>37<\/td>\n<td>96<\/td>\n<td>33,855<\/td>\n<\/tr>\n<tr>\n<td>161.<\/td>\n<td>Thailand<\/td>\n<td>809<\/td>\n<td>3,378<\/td>\n<td>1,151,445<\/td>\n<\/tr>\n<tr>\n<td>162.<\/td>\n<td>Timor-Leste<\/td>\n<td>11<\/td>\n<td>57<\/td>\n<td>9,244<\/td>\n<\/tr>\n<tr>\n<td>163.<\/td>\n<td>Togo<\/td>\n<td>1<\/td>\n<td>1<\/td>\n<td>694<\/td>\n<\/tr>\n<tr>\n<td>164.<\/td>\n<td>Tunisia<\/td>\n<td>13<\/td>\n<td>73<\/td>\n<td>16,137<\/td>\n<\/tr>\n<tr>\n<td>165.<\/td>\n<td>Turkey<\/td>\n<td>631<\/td>\n<td>3,990<\/td>\n<td>2,102,011<\/td>\n<\/tr>\n<tr>\n<td>166.<\/td>\n<td>Uganda<\/td>\n<td>7<\/td>\n<td>17<\/td>\n<td>3,342<\/td>\n<\/tr>\n<tr>\n<td>167.<\/td>\n<td>Ukraine<\/td>\n<td>337<\/td>\n<td>4,731<\/td>\n<td>2,499,272<\/td>\n<\/tr>\n<tr>\n<td>168.<\/td>\n<td>United Arab Emirates<\/td>\n<td>336<\/td>\n<td>1,600<\/td>\n<td>375,077<\/td>\n<\/tr>\n<tr>\n<td>169.<\/td>\n<td>United Kingdom<\/td>\n<td>3,992<\/td>\n<td>142,487<\/td>\n<td>24,480,273<\/td>\n<\/tr>\n<tr>\n<td>170.<\/td>\n<td>United States<\/td>\n<td>6,467<\/td>\n<td>266,984<\/td>\n<td>54,830,162<\/td>\n<\/tr>\n<tr>\n<td>171.<\/td>\n<td>Uruguay<\/td>\n<td>6<\/td>\n<td>13<\/td>\n<td>3,681<\/td>\n<\/tr>\n<tr>\n<td>172.<\/td>\n<td>Uzbekistan<\/td>\n<td>8<\/td>\n<td>19<\/td>\n<td>573<\/td>\n<\/tr>\n<tr>\n<td>173.<\/td>\n<td>Vatican City<\/td>\n<td>6<\/td>\n<td>18<\/td>\n<td>2,570<\/td>\n<\/tr>\n<tr>\n<td>174.<\/td>\n<td>Venezuela<\/td>\n<td>3<\/td>\n<td>5<\/td>\n<td>461<\/td>\n<\/tr>\n<tr>\n<td>175.<\/td>\n<td>Vietnam<\/td>\n<td>347<\/td>\n<td>2,903<\/td>\n<td>465,490<\/td>\n<\/tr>\n<tr>\n<td>176.<\/td>\n<td>Zambia<\/td>\n<td>8<\/td>\n<td>42<\/td>\n<td>4,609<\/td>\n<\/tr>\n<tr>\n<td>177.<\/td>\n<td>Zimbabwe<\/td>\n<td>1<\/td>\n<td>2<\/td>\n<td>5<\/td>\n<\/tr>\n<tr>\n<td class=\"naglasi\" colspan=\"2\">TOTAL<\/td>\n<td class=\"naglasi\">1,368,702<\/td>\n<td class=\"naglasi\">26,701,365<\/td>\n<td class=\"naglasi\">7,235,096,012<\/td>\n<\/tr>\n<tr>\n<td class=\"ukupno\" colspan=\"5\">Last update: Mon Apr 1 08:19:41 CEST 2019<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>Prema dostupnim <a href=\"https:\/\/dev.maxmind.com\/geoip\">MaxMindovim GeoIP<\/a> podatcima, hrvatska vr\u0161na domena raspola\u017ee s ukupno 2.818.597 IP-adresa, od kojih dobar dio nije izravno dostupan krajnjim korisnicima interneta. Prema podatcima iz gornjega prikaza proizlazi da je 41 % hrvatskih IP-adresa koristilo Ha\u0161ekovu uslugu, iz \u010dega slijedi da je on nedvojbeno infrastrukturna usluga u Hrvatskoj. Uzimaju\u0107i u obzir udio Hrvata u populaciji BiH te \u010dinjenicu da je 13 % bosanskohercegova\u010dkih IP-adresa koristilo istu uslugu, zaklju\u010dak se mo\u017ee protegnuti i na tu zemlju. Specifi\u010dnost Ha\u0161eka kao hrvatske infrastrukturne usluge jest ta da nikada nikakve veze nije imao, unato\u010d svim nastojanjima da se takav status promijeni, sa zadu\u017eenima za skrb o nacionalnim interesima. Izvjesno je da to tako ne mo\u017ee i\u0107i do u nedogled, ako ni radi \u010dega drugoga onda radi smrtnosti njegova odr\u017eavatelja.<\/p>\n<h3 class=\"naslov\">Bilje\u0161ke<\/h3>\n<p class=\"biljeska\"><a href=\"#_ftnref1\" name=\"_ftn1\">[1]<\/a> Bentley, J.: A Spelling Checker, <em>Communications of the ACM<\/em>, 28(5), 1985., str. 460.<\/p>\n<p class=\"biljeska\"><a href=\"#_ftnref2\" name=\"_ftn2\">[2]<\/a> Dembitz, \u0160.: Funkcionalna leksikografija mre\u017enoga pravopisnog provjernika, <em>Filologija<\/em>, 58(2012), str. 55-98, HAZU, 2012.<\/p>\n<p class=\"biljeska\"><a href=\"#_ftnref3\" name=\"_ftn3\">[3]<\/a> Uklju\u010duje i promet iniciran iz Republike Kosovo. Premda je po <a href=\"https:\/\/hr.wikipedia.org\/wiki\/ISO_3166-1\">ISO-3166-1<\/a> standardu Kosovu ve\u0107 dodijeljena vr\u0161na domena KO, razdvajanje vr\u0161nih domena Kosova i Srbije jo\u0161 nije obavljeno.<\/p>\n<p class=\"biljeska\"><a href=\"#_ftnref4\" name=\"_ftn4\">[4]<\/a> Citiranu poruku je 27. sije\u010dnja 2019. Ha\u0161eku (<a href=\"mailto:hascheck@fer.hr\">hascheck@fer.hr<\/a>) uputila Tena \u0106ori\u0107, osoba ro\u0111ena i odrasla u \u0160vicarskoj.<\/p>\n<p class=\"biljeska\"><a href=\"#_ftnref5\" name=\"_ftn5\">[5]<\/a> Michel, J.-B., et al.: Quantitative Analysis of Culture Using Millions of Digitized Books, <em>Science<\/em>, Vol. 331, Issue 6014, pp. 176-182, 2011.<\/p>\n<p class=\"biljeska\"><a href=\"#_ftnref6\" name=\"_ftn6\">[6]<\/a> \u0160oi\u0107, R.: <em>Sinteza hrvatskog govora uporabom sustava Festival<\/em>, diplomski rad br. 74, FER, Zagreb, 2010.<\/p>\n<p class=\"biljeska\"><a href=\"#_ftnref7\" name=\"_ftn7\">[7]<\/a> Bajo, D., Turkovi\u0107, D., Dembitz, \u0160.: Rapid Prototyping of a Croatian Large Vocabulary Continuous Speech Recognition System, <em>Proceedings of the IARIA<\/em>, pp. 13-18, Curran Associates, Red Hook, NY, 2014.<\/p>\n<p class=\"biljeska\"><a href=\"#_ftnref8\" name=\"_ftn8\">[8]<\/a> Dembitz, \u0160.: <em>Automatizacija postupka otkrivanja gre\u0161aka u tekstu u novim telekomunikacijskim slu\u017ebama<\/em>, doktorska disertacija, ETF-Zagreb, 1993., str. 5.<\/p>\n<\/div>\n","protected":false},"excerpt":{"rendered":"<p>Donosimo izvorni tekst \u010dlanka o Ha\u0161ekovih 25 godina rada koji je objavljen u \u010dasopisu &#8220;Jezik&#8221;, god. 66, br. 4-5, str. 138-150. Rad je primljen 2. travnja 2019., prihva\u0107en za tisak 7. listopada 2019. i nakon tiskanja pretvoren u ovaj oblik s dopu\u0161tenjem uredni\u0161tva Jezika.<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_exactmetrics_skip_tracking":false,"_exactmetrics_sitenote_active":false,"_exactmetrics_sitenote_note":"","_exactmetrics_sitenote_category":0,"footnotes":""},"categories":[7],"tags":[],"class_list":["post-712","post","type-post","status-publish","format-standard","hentry","category-blog"],"_links":{"self":[{"href":"https:\/\/ispravi.me\/info\/wp-json\/wp\/v2\/posts\/712","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/ispravi.me\/info\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/ispravi.me\/info\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/ispravi.me\/info\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/ispravi.me\/info\/wp-json\/wp\/v2\/comments?post=712"}],"version-history":[{"count":7,"href":"https:\/\/ispravi.me\/info\/wp-json\/wp\/v2\/posts\/712\/revisions"}],"predecessor-version":[{"id":972,"href":"https:\/\/ispravi.me\/info\/wp-json\/wp\/v2\/posts\/712\/revisions\/972"}],"wp:attachment":[{"href":"https:\/\/ispravi.me\/info\/wp-json\/wp\/v2\/media?parent=712"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/ispravi.me\/info\/wp-json\/wp\/v2\/categories?post=712"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/ispravi.me\/info\/wp-json\/wp\/v2\/tags?post=712"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}