Home
BackgroundAgendaLocalesVenueParticipantsWorking GroupResourcesOrganisersPhotosFAQ |
WebSiteTaggingTagging and Machine Translation This project makes existing English/French web content accessible to people with low to no English literacy with having to adapt existing content or websites. Aim is to create tool and infrastructure for collaborative management of the highlighting of difficult English words on website. Then to build Machine Translations that focus on close languages, expanding to languages that are further adaptable. With this all being browser based to prevent the need for centralised infrastructure. Languages: Tagging – South African languages (but generic infrastructure. MACHINE TRANSLATION, Nguni group, then Sotho-Nguni, then English – Nguni Region : South Africa Beneficiaries: Stem Southern African Internet Users, Long term other languages and groups. Achievability: Tagging Yes, Machine Translation close languages – Yes, harder as we get further apart. Risks: Machine Translation as we get broader we reach borders of our knowledge of Machine Translation and the languages. We have skilled people to do Machine Translation (Access of Corpora) Time: Tagging 1 year, Machine Translation, 3 years Skills: Programmer, linguist, Computational linguist translators < Database of African language resources | Off the Wall Projects | Network of African language resources > |