Computing Focus Time of Paragraph Using Deep Learning

Time is a main part of any text that shows value of the document. We have two type of documents that are temporal and atemporal characters. In this paper we explain problem of focus time in historical documents using convolutional neural network (CNN) which shows the time period of text contents and it measure out the document creation time. We used to different type of documents that one is Wikipedia documents and another is web news texts. We evaluate the advantages of our method on various text documents that related to historical events in defined timelines. We used different methods for evaluating focus time of paragraph. To find the relationship between words in document we used different simulation methods and word 2vec system to get more detail results of words. Our proposed approach contains focus time of paragraphs in text documents using deep learning process.