Documents topic classification model in social networks using classifiers voting system
Hyeoncheol Lee, Beomseok Hong, Kwangmi Ko Kim · 2015
Topic model uncovers abstract topics within texts documents, which is an essential task in text analysis in social networks. However, identifying topics in text documents in social networks is challenging since the texts are short, unlabeled, and unstructured. For this reason, we propose a topic classification system regarding the features of text documents in social networks. The proposed system is based on several machine-learning algorithms and voting system. The accuracy of the system has been tested using text documents that were classified into three topics. The experiment results show that the proposed system guarantees high accuracy rates in documents topic classification.