Semantic recognition of words in a sentence using data mining methods and artificial neural networks

In order to convert a sentence from a language into another language, it is first necessary to have a word recognition or word processing procedure and a structural analysis, so that the simple and compound words of that language are recognized by the input. Then the combination of words must be syntactically correct and create the sentence belonging to that language. In this Article, the goal is to create a model for recognizing the semantic role of words in the sentence. To this end, data from the Bijan Khan database has been used. This standard set has a variety of word roles. The model proposed in this research includes all roles. One of the disadvantages of past practices is undesirable perfor-mance in dealing with unknown words. In the next step, this research has tried to use the normalization of the database so that data can be well integrated into the learning process. This is done by data mining software to be done in a logical way. To determine the role of words in the sentence of the sentence structure, the word structure and the database of data words are used.An algorithm used for data mining is neural networks. The neural network used in this re-search uses the famous back propagation algorithm. This algorithm has a high ability to learn. The results indicate a very good accuracy of the deep learning network; the detection rate in this method increased by about 1.5%

Keywords:

Bijan Khan s architecture , artificial neural network , deep learning