Titre : | Machine Translating of Moroccan Dialect | Type de document : | projet fin études | Auteurs : | Maryam OUHAMI, Auteur | Langues : | Français (fre) | Catégories : | SDBD
| Mots-clĂ©s : | NLP, Morphology analyzer, Moroccan Dialect | Index. dĂ©cimale : | mast 93/18 | RĂ©sumĂ© : | The rise of the internet as well as the continued growth of access through different technologies around the world caused an increasing quantities of web content, more specifically the content writing in Arabic, as a result of people’s tendency to express themselves in their own languages. The Arabic content comprise more than 22 types of Arabic dialects within countries and regions. Our main goal is to provide a fair understanding of the Moroccan Arabic web content to the Arabic users, this goal can be achieved using machine translation techniques. For this purpose, a parallel corpus is needed in addition to natural language processing tools.
Since Moroccan Dialect (MD) is not enough studied in the natural language processing area most of MSA tools are not adapted to dialects and do not take into account, and as far as we are concerned no Moroccan parallel corpus is for open sourced. We attempt machine translation of MD within a direct rule-based approach based on deep morphology analysis, besides we built our Moroccan Dialect Parallel Corpus (MDPC) from translating manually an MSA dictionary and some of the MD web content. The findings are interesting, the translation is accomplished at a morphological level.
|
Machine Translating of Moroccan Dialect [projet fin études] / Maryam OUHAMI, Auteur . - [s.d.]. Langues : Français ( fre) Catégories : | SDBD
| Mots-clĂ©s : | NLP, Morphology analyzer, Moroccan Dialect | Index. dĂ©cimale : | mast 93/18 | RĂ©sumĂ© : | The rise of the internet as well as the continued growth of access through different technologies around the world caused an increasing quantities of web content, more specifically the content writing in Arabic, as a result of people’s tendency to express themselves in their own languages. The Arabic content comprise more than 22 types of Arabic dialects within countries and regions. Our main goal is to provide a fair understanding of the Moroccan Arabic web content to the Arabic users, this goal can be achieved using machine translation techniques. For this purpose, a parallel corpus is needed in addition to natural language processing tools.
Since Moroccan Dialect (MD) is not enough studied in the natural language processing area most of MSA tools are not adapted to dialects and do not take into account, and as far as we are concerned no Moroccan parallel corpus is for open sourced. We attempt machine translation of MD within a direct rule-based approach based on deep morphology analysis, besides we built our Moroccan Dialect Parallel Corpus (MDPC) from translating manually an MSA dictionary and some of the MD web content. The findings are interesting, the translation is accomplished at a morphological level.
|
|