Mokgonyane, TBSefara, Tshephisho JManamela, MJModipa, TI2019-12-182019-12-182019-08Mokgonyane, T.B., Sefara, T.J., Manamela, M.J. & Modipa, T.I. 2019. The effects of data size on text-independent automatic speaker identification system. In: International Conference on Advances in Big Data, Computing and Data Communication Systems (icABCD), Winterton, South Africa, 5-6 August 2019978-1-5386-9236-3978-1-5386-9237-0DOI: 10.1109/ICABCD.2019.8851018https://ieeexplore.ieee.org/document/8851018http://hdl.handle.net/10204/11274Presented in: International Conference on Advances in Big Data, Computing and Data Communication Systems (icABCD), Winterton, South Africa, 5-6 August 2019. This is the accepted version of the published item.Speaker recognition is a technique that automatically identifies a speaker from a recording of their speech utterance. Speaker recognition technologies are taking a new direction due to rapid progress in artificial intelligence. Research in the field of speaker recognition has shown fruitful results. There is, however, not much work done for African indigenous languages that have limited speech data resources. This paper presents how data size impacts the accuracy of an automatic speaker recognition system models, focusing on the Sepedi language as it is one of the South African under-resourced language. The speech data used is acquired from the South African Centre for Digital Language Resources. Four machine learning models, namely, Support Vector Machines (SVM), K-Nearest Neighbors (KNN), Multilayer Perceptrons (MLP) and Logistic Regression (LR) are trained under four data setting environment. LR performed better than other models with the highest accuracy of 91% while SVM obtained the highest increase of 4% in accuracy as data size increases.enK-nearest neighborsLogistic regressionMultilayer-perceptronSpeaker recognitionSupport vector machineText-independentConference PresentationMokgonyane, T., Sefara, T. J., Manamela, M., & Modipa, T. (2019). The effects of data size on text-independent automatic speaker identification system. IEEE. http://hdl.handle.net/10204/11274Mokgonyane, TB, Tshephisho J Sefara, MJ Manamela, and TI Modipa. "The effects of data size on text-independent automatic speaker identification system." (2019): http://hdl.handle.net/10204/11274Mokgonyane T, Sefara TJ, Manamela M, Modipa T, The effects of data size on text-independent automatic speaker identification system; IEEE; 2019. http://hdl.handle.net/10204/11274 .TY - Conference Presentation AU - Mokgonyane, TB AU - Sefara, Tshephisho J AU - Manamela, MJ AU - Modipa, TI AB - Speaker recognition is a technique that automatically identifies a speaker from a recording of their speech utterance. Speaker recognition technologies are taking a new direction due to rapid progress in artificial intelligence. Research in the field of speaker recognition has shown fruitful results. There is, however, not much work done for African indigenous languages that have limited speech data resources. This paper presents how data size impacts the accuracy of an automatic speaker recognition system models, focusing on the Sepedi language as it is one of the South African under-resourced language. The speech data used is acquired from the South African Centre for Digital Language Resources. Four machine learning models, namely, Support Vector Machines (SVM), K-Nearest Neighbors (KNN), Multilayer Perceptrons (MLP) and Logistic Regression (LR) are trained under four data setting environment. LR performed better than other models with the highest accuracy of 91% while SVM obtained the highest increase of 4% in accuracy as data size increases. DA - 2019-08 DB - ResearchSpace DP - CSIR KW - K-nearest neighbors KW - Logistic regression KW - Multilayer-perceptron KW - Speaker recognition KW - Support vector machine KW - Text-independent LK - https://researchspace.csir.co.za PY - 2019 SM - 978-1-5386-9236-3 SM - 978-1-5386-9237-0 SM - DOI: 10.1109/ICABCD.2019.8851018 T1 - The effects of data size on text-independent automatic speaker identification system TI - The effects of data size on text-independent automatic speaker identification system UR - http://hdl.handle.net/10204/11274 ER -