Classification with Multiple Classes using Naïve Bayes and Text Generation with a Small Data Set using a Recurrent Neural Network
Abstract
In this thesis, text classification and text generation are explored using only a small
data set and many classes. This thesis experiments with text classification, and show
how it is able to find the most similar output compared to the input even with thousands
of classes. Furthermore, text generation is explored on a small data set to create
a unique output. By using Na¨ıve Bayes text classifier combined with a Recurrent Neural
Network language-model, it is possible to use new deviations as input before an
original suggestion for a measure is generated as the output
Description
Master's thesis Information- and communication technology IKT590 - University of Agder 2017