Mampaka Lydia Mojapelo
The African Wordnet Project (AWN) includes all nine indigenous South African languages, namely isiZulu, isiXhosa, Setswana, Sesotho sa Leboa, Tshivenda, Siswati, Sesotho, isiNdebele and Xitsonga. The AWN currently includes 61 000 synsets as well as definitions and usage examples for a large part of the synsets. The project recently received extended funding from the South African Centre for Digital Language Resources (SADiLaR) and aims to update all aspects of the current resource, including the seed list used for new development, software tools used and mapping the AWN to the latest version of PWN 3.1. As with any resource development project, it is essential to also include phases of focused quality assurance and updating of the basis on which the resource is built. The African languages remain under-resourced. This paper describes progress made in the development of the AWN as well as recent technical improvements.
African WordNet: A Viable Tool for Sense Discrimination in the Indigenous African Languages of South Africa
Stanley Madonsela | Mampaka Lydia Mojapelo | Rose Masubelele | James Mafela
Proceedings of the 8th Global WordNet Conference (GWC)
In promoting a multilingual South Africa, the government is encouraging people to speak more than one language. In order to comply with this initiative, people choose to learn the languages which they do not speak as home language. The African languages are mostly chosen because they are spoken by the majority of the country’s population. Most words in these languages have many possible senses. This phenomenon tends to pose problems to people who want to learn these languages. This article argues that the African WordNet may the best tool to address the problem of sense discrimination. The focus of the argument will be on the primary sense of the word ‘hand’, which is part of the body, as lexicalized in three indigenous languages spoken in South Africa, namely, Tshivenḓa, Sesotho sa Leboa and isiZulu. A brief historical background of the African WordNet will be provided, followed by the definition of the word ‘hand’ in the three languages and the analysis of the word in context. Lastly, the primary sense of the word ‘hand’ across the three languages will be discussed.
This paper presents a linguistic account of the lexical semantics of body parts in African WordNet, with special reference to Northern Sotho. It focuses on external human body parts synsets in Northern Sotho. The paper seeks to support the effectiveness of African WordNet as a resource for services such as in the healthcare and medical field in South Africa. It transpired from this exploration that there is either a one-to-one correspondence or some form of misalignment of lexicalisation with regard to the sample of examined synsets. The paper concludes by making suggestions on how African WordNet can deal with such semantic misalignments in order to improve its efficiency as a resource for the targeted purpose.