Scholarly Works
Permanent URI for this collectionhttps://repository.ui.edu.ng/handle/123456789/347
Browse
3 results
Search Results
Item Rule-Based Machine Translation: An Interface between Formal and Natural Language Syntax A Violation of Case Filter Principle(West African Linguistics Society, 2016) Odoje, C.O.The principles which govern ways words can be combined together to form phrases and sentences in natural language is known as syntax while formal syntax is not a matter of experience (unlike natural language), but stipulations in order to provide a specified set of strings in a computer programming language. The focus of this paper therefore, is to explore linguistics as the dual planes of theory and practice, by interrogating how PROLOG was used to capture English/Yoruba natural language syntax in a rule-based machine translation. The study reveals that the machine was able to generate sentences, break sentences into phrases and words in a bid to translate them in both languagesItem The Peculiar Challenges of SMT to African Languages(The Linguistic Association of Nigeria, 2013) Odoje, C.O.The challenges of Machine translation (MT) have been identified and classified but the classification did not consider the peculiarities of African languages. This paper therefore explores the challenges of MT and reclassifies them in relation with the uniqueness of African languages. The study uses the Yoruba language as a template for other African languages and identifies some of the peculiarities of African languages which include Tie fact that they are resource-scarce languages; dycritization, demarcation of discipline, and funding, amongst others. The paper recommends measures to overcome some of these challenges.Item Towards Machine Translation for Security Surveillance(Faculty of Arts, University of Benin., 2021) Odoje, C. O.; Nweya, G. O.Global insecurity is one of the main challenges facing the world in recent times and Nigeria is among the most affected with thousands of deaths and loss of property worth billions of naira. Studies show that countries are achieving better security through the use intelligence reports where languages play significant roles than through the use of arms and ammunitions. However, previous studies on Nigerian languages have concentrated on language description and language documentation with little attention paid to language use for security surveillance and intelligence gathering This paper, therefore, evaluates Google translate, from the perspective of the Igbo and Yoruba languages. with a view to determining its level of efficiency in translating for the purpose of security surveillance or intelligence gathering, identifying its potentials for achieving better security and the challenges facing its use. The study reveals that Google translate could be used for security surveillance if properly adapted despite the shortcomings of its output. Factors such as multilingualism, inadequate funding, insufficient language resources and poor infrastructural development are some of the challenges facing the proposal. The implication is that the Nigerian government at all levels can harness the potentials of this tool towards overcoming its security challenges if it invests more in security especially at the State and Local Government levels.