Analiza leksičke baze usluge strojnog provjernika pravopisa

undergraduate thesis

• Title (english): Lexical database analysis of an online spellchecker
• Author: Ivan Kordić
• Mentor: Gordan Gledec (mentor)
• Committee member: Gordan Gledec (predsjednik povjerenstva)
• Committee member: Bartol Boras (član povjerenstva)
• Committee member: Kristijan Poje (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2024-09-02, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Analizira se leksička baza usluge ispravi.me, sustava za računalnu provjeru pravopisa koji se desetljećima razvija na Fakultetu elektrotehnike i računarstva u Zagrebu. Leksička baza usluge temelji se na n-gramskom sustavu. Kako bi se razvila mogućnost ispravljanja ne samo pravopisnih, već i gramatičkih pogrešaka, potrebno je iz n-gramske baze stvoriti bazu koja ne sadrži samo n-grame i njihove frekvencije, nego i gramatičke kategorije riječi. Ovim je radom u velikom postotku pokriven općejezični fond riječi leksičke baze usluge uporabom javno dostupnog leksikona hrLex koji sadrži gramatičke kategorije riječi. Napravljen je algoritam za jednostavnu gramatičku analizu teksta te se razmatra razvoj statističkog modela za označavanje gramatičkih kategorija riječi po uzoru na skrivene Markovljeve lance i uvjetna slučajna polja (CRF).

Abstract (english)

The subject of analysis is the lexical database of the internet service ispravi.me, a system for spell-checking texts that is being developed at the Faculty of Electrical Engineering and Computing in Zagreb. The lexical database of the service is based on the n-gram system. In order to develop the possibility of correcting not only spelling, but also grammatical errors, it is necessary to create a database from the n-gram database that contains not only n-grams and their frequency, but also grammatical categories of words. A large portion of words from the database is covered using the public available lexicon hrLex which contains grammatical categories of words. An algorithm for simple grammar analysis is implemented and the development of a statistical model for marking grammatical categories of words based on hidden Markov chains and Conditional random fields in considered.

• Keywords: obrada prirodnog jezika; gramatička analiza teksta; sintaksna analiza; n-grami; statistički modeli
• Keywords (english): natural language processing; grammatical text analysis; syntax analysis; n-grams; statistical models
• Language: croatian
• URN:NBN: urn:nbn:hr:168:304453
• Study programme: Title: Computing; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2024-10-12 21:06:57