A mirror neuron based neural network for website classification

Mirror neurons become responsive whenever an animal performs a peculiar action or when it observes a certain action being performed. The discovery of these neurons in humans has explained quite a lot of our behaviour patterns. The decoding logic of these neurons forms the vital component to the proposed neural network in this paper. The backdrop to the central idea of mirror neurons is the notion that people's behaviour tends to effect the way in which they browse the web. Using these ideas a neural network can be built based on the working of mirror neurons which can play a vital role in classifying websites into categories. DMOZ provides an open directory of websites and using this large database as a training set, the neural network is trained. The neural network is then subjected to websites that are not listed in DMOZ and are classified based on their mirroring percentage. The decoding principles of mirror neurons are discussed and their role in the foundation for the design of the neural network is explained in detail. The results obtained can be used to effectively classify new websites into one of the DMOZ categories.