Dynamic and adaptive self organizing maps applied to high dimensional large scale text clustering

The self organizing maps(SOM) has been used as a tool for mapping high-dimensional input data into a low-dimensional feature map, which has significant advantages for text clustering applications. In this paper, a novel dynamic and adaptive SOM algorithm applied to high dimensional large scale text clustering is proposed. The characteristic feature of this novel neural network model is its dynamic architecture which grows (when the similarity between input pattern (text vector) and weight vector of the winning node is smaller than a given threshold) during its training process to find the inherent topology structure of the document set. By using unsupervised competitive learning in network, the weight vectors of the winning node and its nearest neighbors are adjusted adaptively (where learning rate is related to similarity in amended learning rule) in this algorithm. The results of the experiments indicated that the algorithm successfully improve quality of text clustering and learning speed of neural network.