Study on the Classification of Mixed Text Based on Conceptual Vector Space Model and Bayes

Traditional vector-space-based text-classification models are established by calculating the weights of feature words on the lexical level. In such models, words are independent on one another and their semantic relations are unrevealed. This paper proposes a vector-space-based text analyzer by introducing conceptual semantic similarity into traditional vector-space-based models. Naive Bayes classification technology is also adopted into this new analyzer. Experiment results indicate that the new analyzer can improve text classification.