Bibliša: Aligned Collection Search Tool

[ Log In ] [ Register ]

 [About]



En/De/Fr/It- (first 9 out of 17 sentences) [pdf] Srpski - (prvih 9 od 17 rečenica) [pdf]
n1Dr. Anna Jouravel, Martin Meindl, Prof. Dr. Achim Rabus, Elena Renjen1Dr. Anna Jouravel, Martin Meindl, Prof. Dr. Achim Rabus, Elena Renje
n2University of Freiburg n2University of Freiburg
n3{anna.jouravel, martin.meindl, achim.rabus, elena.renje}@slavistik.uni-freiburg.de n3{anna.jouravel, martin.meindl, achim.rabus, elena.renje}@slavistik.uni-freiburg.de
n4CREATING A MULTI-LAYERED DIGITAL FRAMEWORK FOR ANALYSING CHURCH SLAVONIC MANUSCRIPTSn4KREIRANJE MULTISTRATALNOG DIGITALNOG OKRUŽENJA ZA ANALIZU CRKVENOSLOVENSKIH RUKOPISA
n5In this contribution, we will discuss various digital methods for analysing Old Church Slavonic as well as Church Slavonic texts in their East and South Slavonic recensions.n5U ovom izlaganju govorićemo o različitim digitalnim metodama za analizu staroslovenskih i crkvenoslovenskih tekstova istočnoslovenske i južnoslovenske redakcije.
n6 For the analysis, we employ a multi-layered approach using various tools and methods rather than focusing on a single approach.n6 Za analizu koristimo multistratalni, a ne monostratalni pristup, uz primenu različitih alata i metoda.
n7 We believe this strategy will yield more valid and robust results, as it is less likely that all the methods combined will be affected by the errors present in the data compared to using one single tool. n7 Verujemo da ovakva strategija daje validnije i robusnije rezultate, jer je manje verovatno da će skup kombinovanih metoda sadržati one greške koje prate primenu samo jednog alata.
n8These methods are part of the quantitative pipeline that begins with the automatic transcription of (Old) Church Slavonic manuscripts and continues with the analysis of these transcriptions, without the need for manual correction of the errors produced by the Handwritten Text Recognition (HTR) engine.n8 Ove metode deo su kvantitativnog lanca koji počinje automatskom transkripcijom (staro)crkvenoslovenskih rukopisa, a nastavlja se analizom ovih transkripata, bez potrebe za manuelnim ispravljanjem grešaka koje nastaju primenom programa za prepoznavanje teksta napisanog rukom (HTR).
n9 This pipeline comprises multiple pre- and postprocessing steps, including automatic transcription, tokenization, parsing, tagging and statistical analysis.n9 Kvantitativni lanac obuhvata i nekoliko koraka koji prethode obradi ili joj slede, uključujući automatsku transkripciju, tokenizaciju, raščlanjivanje, anotaciju (tagiranje) i statističku analizu.