SADiLaR Language Resource Repository
Not a member yet
536 research outputs found
Sort by
DSAE Print Citations Database
No full textC . 300 000 index cards collected from 1969 onwards to document English words with a unique meaning / usage in South Africa, as research process for A Dictionary of South African English on Historical Principles (http://dsae.co.za)
IsiXhosa multi-speaker TTS corpus
No full textThe aim of this corpus was to investigate the implementation of a high-quality TTS system using multiple voices recorded using a low-cost process (i.e. using non-professional volunteers and informal recording environments)
UNISA English/Zulu Parallel Corpus
No full textThe resource comprises sentence aligned and tokenized parallel text in English and Zulu. The text was extracted from the following sources: an adapted version of the English/Zulu Autshumato corpus, paragraph translated Wikipedia texts, the Bible, the Book of Mormon, the Constitution of South Africa, the Universal Declaration of Human Rights and a selection of translated sentences from the book "Beyond the He/Man" (1996)
Qfrency TTS Afrikaans Kobus recordings
No full textStudio quality recordings of text-to-speech data in Afrikaans and some English utterances. Professional Afrikaans first language voice artist
Afrikaans speaking children's first lexical items
No full textData collected for a master's study in Afrikaans linguistics. The data consist of the first lexical items of 21 Afrikaans speaking children. The lexical items were noted in diary format by the children's mothers. For each lexical item, a number of contexts of use are also given as well as other relevant information. The number of lexical items used by each child differs. Pseudonyms were given for the children
Human Language Technology Audit 2017/18
No full textThis document reports on all work conducted in the 2017/18 Audit of human language technology (HLT) resources available in South Africa project. The purpose of conducting the HLT Audit 2017/18 is to update the previous HLT Audit conducted by Sharma Grover in 2009 as part of her research thesis. Increased activity in the HLT field in South Africa, an increase in the number of institutions undertaking HLT research and development (R&D) in the country, the increased local interest in digital humanities as a field of study and the recent establishment of the South African Centre for Digital Language Resources, form part of the rationale for undertaking this project - eight years after the previous Audit.
The HLT Audit 2017/18 project consists of five work packages encompassing the Audit design, instrument development, execution, and analysis, as well as a work package aimed at designing a dynamic Audit update system which will enable entities involved in HLT R&D to upload the outputs of their work as this becomes available.
The HLT Research Group at the CSIR Meraka Institute collaborated with the South African Centre for Digital Language Resources (SADiLaR) and the Resource Management Agency (RMA) to achieve the desired outcomes of the project.
This final report provides detailed information on the work completed from 1 July 2017 to 31 August 2018
Rhonda
No full textRhonda is a speech to speech translation system that connects individual HLTs together in a secure and accessible way. A user may access individual HLTs or request a composite process such as speech to speech translation. Access to Rhonda is achieved via RESTful endpoints that a client application may consum
CGE's Sesotho Gender Terminology List
No full textCGE's Sesotho Gender Terminology List is a list of terms, either words or phrases, related to the promotion of gender equality. All 446 words or phrases appear in English, with definitions in English, and with equivalent Sesotho words or phrases. The list is the result of a collaborative effort between the Commission for Gender Equality (CGE) in partnership with the PanSALB Sesotho National Language Body (SNLB), one of PanSALB's language boards
Test treebank for the LFG/XLE treebank
No full textA selection of 828 Tswana sentences with their LFG/XLE parse tree
South African Broadcast News (SABN) Corpus
No full textThe corpus consists of approximately 20 hours of audio recordings from one of the country's main radio news channels, SAFM. Bulletins were broadcast between 1996 and 2006 and are a mix of news-reader speech, interviews, and crossings to reporter