Please use this identifier to cite or link to this item:
https://www.um.edu.mt/library/oar/handle/123456789/74843
Title: | Inflection of Maltese using neural networks |
Authors: | Schneeberger, Georg (2019) |
Keywords: | Maltese language -- Inflection Neural networks (Computer science) -- Malta Computational linguistics -- Malta |
Issue Date: | 2019 |
Citation: | Schneeberger, G. (2019). Inflection of Maltese using neural networks (Bachelor's dissertation). |
Abstract: | Digitisation, globalisation and the prevalence of English have lead to increased accessibility for information, education, research and international funds and programs. At fi rst that may sound like a purely positive development, unfortunately many languages, especially languages with few speakers, can get left behind in research. This project will focus on a particularly interesting aspect of the Maltese language, inflection. Recent research in the domain of inflection will be analysed and analysed in the Maltese context. In this research a previous state-of-the-art system for inflection will be built and evaluated on Maltese. The results will show if the neural state-of-the-art encoder-decoder systems are truly able to learn the complex inflection of the Maltese language. The evaluation shows that noun inflections in Maltese are much more difficult to learn for the encoder-decoder system utilised in this task compared to the complete paradigms of Maltese verbs. The project's results could not show whether the encoder-decoder architecture is fit for reliable inflection of the complete paradigms of Maltese verbs, nouns and adjectives. It could neither be proven nor disproven. |
Description: | B.SC.ICT(HONS)ARTIFICIAL INTELLIGENCE |
URI: | https://www.um.edu.mt/library/oar/handle/123456789/74843 |
Appears in Collections: | Dissertations - FacICT - 2019 Dissertations - FacICTAI - 2019 |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
Schneeberger Georg.pdf Restricted Access | 2.46 MB | Adobe PDF | View/Open Request a copy |
Items in OAR@UM are protected by copyright, with all rights reserved, unless otherwise indicated.