Title :
Text Message Authorship Classification Using Kernel Support Vector Machines
Author :
Kretchmar, Matt ; Yifu Zhao
Author_Institution :
Dept. of Math. & Comput. Sci., Denison Univ., Granville, OH, USA
Abstract :
We explore the application of Kernel Support Vector Machines (SVM) to the realm of text messages. Our intent is to classify the author of a text message based on usage patterns present in a training set of text messages. We achieve between 57% and 96% accuracy in determining the author of unknown samples.
Keywords :
electronic messaging; pattern classification; support vector machines; SVM; kernel support vector machines; text message authorship classification; text message training set; usage patterns; Accuracy; Educational institutions; Kernel; Support vector machines; Testing; Training; Vectors; Classification; Machine Learning Applications; NLP; SVM; Text Messages;
Conference_Titel :
Computational Science and Computational Intelligence (CSCI), 2014 International Conference on
Conference_Location :
Las Vegas, NV
DOI :
10.1109/CSCI.2014.121
