Linguists have taught the computer to understand the degree of emotion of the text

© Illustration RIA Novosti . Alina Polyaminoamide linguistic analysis of the text in the representation of the artistLinguists have taught the computer to understand the degree of emotion of the text© Illustration RIA Novosti . Alina Polyanina

The linguists of the Institute of Philology and language communication of Siberian Federal University created a technology for automatic analysis of the strength of emotion contained in the text. The developers note that the project is in its importance is at the international level.Applied linguistics is a special interdisciplinary field that works at the intersection of literature, psychology, neurophysiology, information technology, philosophy. The results of studies of texts and speech find application practically in all spheres of human activity. Development of applied linguistics used by intelligence agencies, psychologists, advertisers and spin doctors.

Existing systems for automatic text analysis allow us to estimate only the positive, the negative coloring of the text based on key words. The development of Krasnoyarsk scientists can instantly «understand» not only the color of the text, but also to assess the emotional experiences contained in it.

Based on a sample of texts which reflected basic emotions (e.g. fear, anger, disgust, joy, surprise, etc.), was selected characteristic describing their words and combinations (markers) and the corresponding linguistic model. The presence of markers in the text allows us to evaluate what emotions the text contains, what it shades of emotional experience, how strong they are.

«A ranked classifier is very popular for the analysis of messages, posts on social networks, — says head of the Department of romance languages and applied linguistics at SFU, doctor of Philology, Professor Anastasia Kolmogorov. The customers may be industrial companies, marketing structure, political parties.»

Another application development SFU, on which the team of scientists working at the present moment, is able to evaluate the texts according to the level of manipulative influence. The created algorithm allows the computer to identify signs of hidden manipulation in the text, the presence of calls for illegal action. A software product may be claimed, when the need for an independent linguistic examination of the text.