Morphology of Network

Word Morphology gets rid of morphemes, instead speaking to all morphology as relations among sets of words, which we call lexical correspondences. This paper shows a more formal treatment of Whole Word Morphology than has been previously distributed, exhibiting how the morphological relations are interceded by unification with sequence variables. We present a system for morphological re-inflection dependent on an encoder-decoder taxonomical network model with additional convolution layers. The hypothesis of artificial taxonomical networks has been successfully connected to a wide assortment of pattern recognition issues. In this hypothesis, the initial phase in computing the following condition of a neuron or in performing the following layer taxonomical network computation involves the linear task of duplicating taxonomical values by their synaptic strengths and including the results.