Using Vector Proximity for NLP Analysis of Specialized Texts
Alexander Sak · 2022
The paper considers some latest approaches to develop a machine translation system to handle texts specialized in construction industry both in Russian and English. Graph theory is the basis for automatic sentence analysis and representation of its structure. At the first stage, the sentence is represented by a line graph. The article discusses how to assign attributes to each word in accordance with its prospective belonging to different parts of speech for their subsequent processing using graph convolutional neural networks (GCN). The initial features of each element of the sentence are compared with the calculated values obtained, and by means of certain parameters for the compiled vectors, they are classified as parts of speech. The paper contains parts of the software listing for the computer implementation of the proposed technology.