PROPOSAL CLASSIFICATION ALGORITHM OF VIETNAMESE TEXT USING LONG SHORT TERM MEMORY AND WORD2VECABSTRACT
139 viewsKeywords:
Text Classification; Natural Language Processing; Long Short Term Memory; Word2vec; Data Processing.Abstract
Recently, text classification is considered as a fundamental approach in Natural Language Processing (NLP). It can be widely applied into numerous fields namely sentiment analyses, topic labelings and so on. Specifically, recent achievements have shown that Deep Learning (DL) methods obtained great performance in classifying texts. These methods have positive effects on text classification, especially in English. However, there are few studies investigating about their impacts on Vietnamese text classification. Therefore, in this research, Long Short Term Memory (LSTM) network and Word2Vec engine were used in text classification with the aim of improving efficiency and accuracy. The results of model evaluation on Vietnamese text VNTC [1] we concluded were feasible and likely to be applied in real-life contexts in the near future.
