search
The following results are related to Rural Digital Europe. Are you interested to view more results? Visit OpenAIRE - Explore.

  • Rural Digital Europe
  • Doctoral thesis
  • Croatian
  • ODRAZ - Open Repository of the Univ...

Relevance
arrow_drop_down
  • image/svg+xml Jakob Voss, based on art designer at PLoS, modified by Wikipedia users Nina and Beao Closed Access logo, derived from PLoS Open Access logo. This version with transparent background. http://commons.wikimedia.org/wiki/File:Closed_Access_logo_transparent.svg Jakob Voss, based on art designer at PLoS, modified by Wikipedia users Nina and Beao
    Authors: Trbušić, Željko;

    The goal of the Ph.D. thesis titled Methods for Analysis and Process Optimisation of Optical Character Recognition in Archival Information Systems is to investigate the optical character recognition (OCR) technology and its application in the context of the development and upgrading of archival information systems. The study of OCR technology belongs to the field of computer vision, artificial intelligence, and pattern recognition because it enables computers, machines, and robots to perceive their surroundings and use the collected data in their operative environment. Archives deal with a lot of textual data and the use of OCR delivers many benefits to the management of archival records. It is, therefore, necessary to build a model that can enable its trustworthiness, long-term sustainability, and optimisation. The development of such methodology begins with the research of the working environment and the variables that surround the process of optical character recognition and that can have a great impact on the implementation procedure. In this context, this thesis delivers historical information about the development of OCR, but also it describes the modern use of OCR which is centred around cloud-based solutions and artificial intelligence. The variables presented are concerned with the technological surroundings of the OCR process: the archival materials which are the recognition subject, typography which plays a central role in the ability to “read” the printed text, devices which are used in the digitization process, image file formats and compression standards, and text file formats and the text encoding standards. The thesis analyses three archival information systems: OAIS reference model (OAIS RM), HathiTrust digital library, and the information system of National Library of Medicine (NLM). In the context of OAIS RM, the method of OCR implementation is delivered on a conceptual level, the HathiTrust system defines the process of recognition on the level of communication with the information producers, and the NLM constructed an in-house OCR process which is analysed in detail. Next, the four phases of OCR implementation process are introduced. They enable the implementation of evaluation procedures that are essential in the further development of archival information systems. The evaluation process, the key to the purposeful and systematic research presented in this thesis, is defined, and its main steps and units of measurement are presented as well as The ISRI Analytic Tools for OCR Evaluation. The last section of the thesis presents the results of the evaluation process that was conducted using archival materials in Croatian and English languages. The testing process was vi organised in three separate operating environments which encompass a broad spectrum of archival materials, and which deliver conclusions that can benefit the optimisation of optical character recognition process in every segment. The research investigates financial and timerelated savings but also addresses the concepts of trustworthiness and information value of the finished product. The evaluation process was conducted using The ISRI Tools and it comprised of fundamental statistical analysis that can be used not only in the evaluation of OCR but is useful in the broader evaluation of archival information systems. This thesis is based on a principle that perfect archival information system or flawlessly implemented OCR process should not be the main goal of archival endeavours (even though the quantitative analysis, which is the main device used throughout the research, could suggest otherwise). The OCR technology is not capable of delivering perfect results of the recognition process, but it can nevertheless be used in the field of management of archival materials if its performance is measured, optimised, and systemised. The results can positively influence the usability of stored archival data and elevate the trust that users have in the archives and the archival community. Doktorska disertacija naslovljena Metode analize i optimizacije procesa optičkog prepoznavanja znakova u arhivskim informacijskim sustavima izrađena je s ciljem istraživanja tehnologije optičkog prepoznavanja znakova (OCR) i njene primjene u okviru izgradnje i unapređenja arhivskih informacijskih sustava. Ova tehnologija pripada u šira područja proučavanja tehnologije računalnog vida, umjetne inteligencije i prepoznavanja uzoraka koja omogućuju računalima, strojevima i robotima da percipiraju analogne informacije iz okoline te ih koriste u vlastitom radu. S obzirom na veliku količinu tekstualnih informacija koje arhivi pohranjuju, potrebno je osmisliti model upravljanja, implementacije i evaluacije optičkog prepoznavanja znakova kojim će se ostvariti korištenje ove tehnologije na vjerodostojan, dugoročno održiv i optimalan način. Izgradnja takvog metodološkog procesa započinje proučavanjem radnog okruženja i vanjskih utjecaja čije zakonitosti određuju mogućnost i način implementacije. Obrađen je povijesni razvoj ove tehnologije koji prikazuje razvoj od samih početaka pa sve do suvremenih sustava koji koriste usluge u oblaku i mehanizme potpomognute umjetnom inteligencijom. Također je pružen pregled svih tehnoloških aspekata koji utječu na proces optičkog prepoznavanja znakova, od analize specifičnosti arhivskoga gradiva koje se prepoznaje, tipografskih elemenata slovnih oznaka, uređaja kojima se obavlja digitalizacija i povezanih formata prijenosa slikovnih datoteka do formata prijenosa računalno kodiranog teksta i načina kodiranja znakova. Obrađen je OAIS referentni model na temelju kojeg su prikazane mogućnosti implementacije optičkog prepoznavanja znakova na konceptualnoj razini, a predstavljena su i dva funkcionalna arhivska informacijska sustava koja koriste OCR u svakodnevnom poslovanju. Sustav zaklade Hathi definira proces prepoznavanja na razini komunikacije sa stvarateljima dok informacijski sustav Nacionalne medicinske knjižnice SAD-a obavlja postupak prepoznavanja izgradnjom unutrašnjih tehnoloških rješenja. Identificirani su elementi faza procesa optičkog prepoznavanja znakova koji omogućuju provođenje evaluacijskih metoda te unapređenje sustava u svim njegovim segmentima. U sklopu procesa evaluacije, ključnog za provođenje svrsishodnih i sistematički značajnih istraživanja prikazanih u ovom radu, definirani su koraci testiranja, mjerne jedinice te su detaljno predstavljeni ISRI analitički alati za evaluaciju OCR-a. iv Posljednji dio rada predstavlja rezultate i analizu evaluacije provedene na arhivskom gradivu na hrvatskom i engleskom jeziku. Testiranja su organizirana u tri radne okoline koje obuhvaćaju širok spektar arhivskoga gradiva i kojima se donose zaključci koji omogućuju optimizaciju procesa optičkog prepoznavanja znakova u svih njegovim fazama. Promatraju se postignute financijske i vremenske uštede, ali također i vjerodostojnost evaluiranog procesa te informacijska vrijednost produkta prepoznavanja. Testiranja, obavljena ISRI sustavom, obuhvatila su i temeljne statističke analize koje se mogu upošljavati ne samo kao dio evaluacije procesa optičkog prepoznavanja, već i za evaluaciju cjelokupnog sustava u svim njegovim segmentima. Ovaj rad polazi iz perspektive koja ne traga za savršenim sustavom niti besprijekorno implementiranim procesom, već se, pomalo paradoksalno za kvantitativnu analizu koja je temelj provedenih eksperimenata, dokazuje kako se i nesavršena tehnologija kao što je OCR može upotrijebiti u području upravljanja arhivskim gradivom na sustavan i optimalan način koji doprinosi povećanju iskoristivosti pohranjenog gradiva i podiže povjerenje koje korisnici imaju u arhive i arhivsku zajednicu.

    image/svg+xml Jakob Voss, based on art designer at PLoS, modified by Wikipedia users Nina and Beao Closed Access logo, derived from PLoS Open Access logo. This version with transparent background. http://commons.wikimedia.org/wiki/File:Closed_Access_logo_transparent.svg Jakob Voss, based on art designer at PLoS, modified by Wikipedia users Nina and Beao ODRAZ - Open Reposit...arrow_drop_down
    image/svg+xml Jakob Voss, based on art designer at PLoS, modified by Wikipedia users Nina and Beao Closed Access logo, derived from PLoS Open Access logo. This version with transparent background. http://commons.wikimedia.org/wiki/File:Closed_Access_logo_transparent.svg Jakob Voss, based on art designer at PLoS, modified by Wikipedia users Nina and Beao
    image/svg+xml Jakob Voss, based on art designer at PLoS, modified by Wikipedia users Nina and Beao Closed Access logo, derived from PLoS Open Access logo. This version with transparent background. http://commons.wikimedia.org/wiki/File:Closed_Access_logo_transparent.svg Jakob Voss, based on art designer at PLoS, modified by Wikipedia users Nina and Beao
    0
    citations0
    popularityAverage
    influenceAverage
    impulseAverage
    BIP!Powered by BIP!
    more_vert
Powered by OpenAIRE graph
search
The following results are related to Rural Digital Europe. Are you interested to view more results? Visit OpenAIRE - Explore.
  • image/svg+xml Jakob Voss, based on art designer at PLoS, modified by Wikipedia users Nina and Beao Closed Access logo, derived from PLoS Open Access logo. This version with transparent background. http://commons.wikimedia.org/wiki/File:Closed_Access_logo_transparent.svg Jakob Voss, based on art designer at PLoS, modified by Wikipedia users Nina and Beao
    Authors: Trbušić, Željko;

    The goal of the Ph.D. thesis titled Methods for Analysis and Process Optimisation of Optical Character Recognition in Archival Information Systems is to investigate the optical character recognition (OCR) technology and its application in the context of the development and upgrading of archival information systems. The study of OCR technology belongs to the field of computer vision, artificial intelligence, and pattern recognition because it enables computers, machines, and robots to perceive their surroundings and use the collected data in their operative environment. Archives deal with a lot of textual data and the use of OCR delivers many benefits to the management of archival records. It is, therefore, necessary to build a model that can enable its trustworthiness, long-term sustainability, and optimisation. The development of such methodology begins with the research of the working environment and the variables that surround the process of optical character recognition and that can have a great impact on the implementation procedure. In this context, this thesis delivers historical information about the development of OCR, but also it describes the modern use of OCR which is centred around cloud-based solutions and artificial intelligence. The variables presented are concerned with the technological surroundings of the OCR process: the archival materials which are the recognition subject, typography which plays a central role in the ability to “read” the printed text, devices which are used in the digitization process, image file formats and compression standards, and text file formats and the text encoding standards. The thesis analyses three archival information systems: OAIS reference model (OAIS RM), HathiTrust digital library, and the information system of National Library of Medicine (NLM). In the context of OAIS RM, the method of OCR implementation is delivered on a conceptual level, the HathiTrust system defines the process of recognition on the level of communication with the information producers, and the NLM constructed an in-house OCR process which is analysed in detail. Next, the four phases of OCR implementation process are introduced. They enable the implementation of evaluation procedures that are essential in the further development of archival information systems. The evaluation process, the key to the purposeful and systematic research presented in this thesis, is defined, and its main steps and units of measurement are presented as well as The ISRI Analytic Tools for OCR Evaluation. The last section of the thesis presents the results of the evaluation process that was conducted using archival materials in Croatian and English languages. The testing process was vi organised in three separate operating environments which encompass a broad spectrum of archival materials, and which deliver conclusions that can benefit the optimisation of optical character recognition process in every segment. The research investigates financial and timerelated savings but also addresses the concepts of trustworthiness and information value of the finished product. The evaluation process was conducted using The ISRI Tools and it comprised of fundamental statistical analysis that can be used not only in the evaluation of OCR but is useful in the broader evaluation of archival information systems. This thesis is based on a principle that perfect archival information system or flawlessly implemented OCR process should not be the main goal of archival endeavours (even though the quantitative analysis, which is the main device used throughout the research, could suggest otherwise). The OCR technology is not capable of delivering perfect results of the recognition process, but it can nevertheless be used in the field of management of archival materials if its performance is measured, optimised, and systemised. The results can positively influence the usability of stored archival data and elevate the trust that users have in the archives and the archival community. Doktorska disertacija naslovljena Metode analize i optimizacije procesa optičkog prepoznavanja znakova u arhivskim informacijskim sustavima izrađena je s ciljem istraživanja tehnologije optičkog prepoznavanja znakova (OCR) i njene primjene u okviru izgradnje i unapređenja arhivskih informacijskih sustava. Ova tehnologija pripada u šira područja proučavanja tehnologije računalnog vida, umjetne inteligencije i prepoznavanja uzoraka koja omogućuju računalima, strojevima i robotima da percipiraju analogne informacije iz okoline te ih koriste u vlastitom radu. S obzirom na veliku količinu tekstualnih informacija koje arhivi pohranjuju, potrebno je osmisliti model upravljanja, implementacije i evaluacije optičkog prepoznavanja znakova kojim će se ostvariti korištenje ove tehnologije na vjerodostojan, dugoročno održiv i optimalan način. Izgradnja takvog metodološkog procesa započinje proučavanjem radnog okruženja i vanjskih utjecaja čije zakonitosti određuju mogućnost i način implementacije. Obrađen je povijesni razvoj ove tehnologije koji prikazuje razvoj od samih početaka pa sve do suvremenih sustava koji koriste usluge u oblaku i mehanizme potpomognute umjetnom inteligencijom. Također je pružen pregled svih tehnoloških aspekata koji utječu na proces optičkog prepoznavanja znakova, od analize specifičnosti arhivskoga gradiva koje se prepoznaje, tipografskih elemenata slovnih oznaka, uređaja kojima se obavlja digitalizacija i povezanih formata prijenosa slikovnih datoteka do formata prijenosa računalno kodiranog teksta i načina kodiranja znakova. Obrađen je OAIS referentni model na temelju kojeg su prikazane mogućnosti implementacije optičkog prepoznavanja znakova na konceptualnoj razini, a predstavljena su i dva funkcionalna arhivska informacijska sustava koja koriste OCR u svakodnevnom poslovanju. Sustav zaklade Hathi definira proces prepoznavanja na razini komunikacije sa stvarateljima dok informacijski sustav Nacionalne medicinske knjižnice SAD-a obavlja postupak prepoznavanja izgradnjom unutrašnjih tehnoloških rješenja. Identificirani su elementi faza procesa optičkog prepoznavanja znakova koji omogućuju provođenje evaluacijskih metoda te unapređenje sustava u svim njegovim segmentima. U sklopu procesa evaluacije, ključnog za provođenje svrsishodnih i sistematički značajnih istraživanja prikazanih u ovom radu, definirani su koraci testiranja, mjerne jedinice te su detaljno predstavljeni ISRI analitički alati za evaluaciju OCR-a. iv Posljednji dio rada predstavlja rezultate i analizu evaluacije provedene na arhivskom gradivu na hrvatskom i engleskom jeziku. Testiranja su organizirana u tri radne okoline koje obuhvaćaju širok spektar arhivskoga gradiva i kojima se donose zaključci koji omogućuju optimizaciju procesa optičkog prepoznavanja znakova u svih njegovim fazama. Promatraju se postignute financijske i vremenske uštede, ali također i vjerodostojnost evaluiranog procesa te informacijska vrijednost produkta prepoznavanja. Testiranja, obavljena ISRI sustavom, obuhvatila su i temeljne statističke analize koje se mogu upošljavati ne samo kao dio evaluacije procesa optičkog prepoznavanja, već i za evaluaciju cjelokupnog sustava u svim njegovim segmentima. Ovaj rad polazi iz perspektive koja ne traga za savršenim sustavom niti besprijekorno implementiranim procesom, već se, pomalo paradoksalno za kvantitativnu analizu koja je temelj provedenih eksperimenata, dokazuje kako se i nesavršena tehnologija kao što je OCR može upotrijebiti u području upravljanja arhivskim gradivom na sustavan i optimalan način koji doprinosi povećanju iskoristivosti pohranjenog gradiva i podiže povjerenje koje korisnici imaju u arhive i arhivsku zajednicu.

    image/svg+xml Jakob Voss, based on art designer at PLoS, modified by Wikipedia users Nina and Beao Closed Access logo, derived from PLoS Open Access logo. This version with transparent background. http://commons.wikimedia.org/wiki/File:Closed_Access_logo_transparent.svg Jakob Voss, based on art designer at PLoS, modified by Wikipedia users Nina and Beao ODRAZ - Open Reposit...arrow_drop_down
    image/svg+xml Jakob Voss, based on art designer at PLoS, modified by Wikipedia users Nina and Beao Closed Access logo, derived from PLoS Open Access logo. This version with transparent background. http://commons.wikimedia.org/wiki/File:Closed_Access_logo_transparent.svg Jakob Voss, based on art designer at PLoS, modified by Wikipedia users Nina and Beao
    image/svg+xml Jakob Voss, based on art designer at PLoS, modified by Wikipedia users Nina and Beao Closed Access logo, derived from PLoS Open Access logo. This version with transparent background. http://commons.wikimedia.org/wiki/File:Closed_Access_logo_transparent.svg Jakob Voss, based on art designer at PLoS, modified by Wikipedia users Nina and Beao
    0
    citations0
    popularityAverage
    influenceAverage
    impulseAverage
    BIP!Powered by BIP!
    more_vert
Powered by OpenAIRE graph