Institute for System Programming of the Russian Academy of Sciences

16 March, 2017

Texterra and Talisman. ISP RAS technologies will change the face of the Internet

Texterra VizOntia"Talisman" and "Texterra" technologies, developed by the Institute for System Programming of the Russian Academy of Sciences, can change the face of the global information network Internet. These products are designed to analyze Big Data of social networks.

"Texterra" is a technology of multilingual text mining. It is based on software developed for accumulating and analyzing information on the concept level and performing intellectual analysis of texts. At the same time, "Texterra" draws on a range of concepts and terms, almost an order of magnitude greater than the size of the British Encyclopedia, and can expand its knowledge base, if necessary. As a result, when using this technology, it is possible to calculate lexically similar texts belonging to a pen of the same author or written according to a common plan, containing similar semantic constructions, and thus, to reveal the fact and structure of information attacks.

The "Talisman" system, in its turn, enables you to track and identify fake accounts that serve for spreading false information, spam, and mass activity simulating - bots. The analysis is based on complete and incomplete data specified in the profiles of social networks, social intercourse of accounts, subscriptions, lists of friends and followers. Values of demographic attributes such as the user age, as well as marital status, education level and so on are detected and verified as a result of the analysis. If the values of attributes do not fit answers produced by the software, we can say confidently that the social network profile falling out of the scheme does not belong to a living person, but is created artificially. Then "Texterra" comes to help, allowing to detect sets of bots created with the same hand or conducted on one method-sheet.

Based on analysis of incomplete data from social networks, the "Talisman" technology allows to identify the approximate primary residence of users, their particular social stratum affiliation, commitment to one or another ideology, and other useful information.

"Analysis of such constantly changing data, like social networks information, is an urgent need today," said Arutyun Avetisyan, Director of ISP RAS, “This is a marketing tool that enables you to evaluate the audience, build patterns that help you carry out target-focused work with different age and social groups. And in many ways, a security tool. In recent decades, unlimited Internet anonymity has become a source of threat for individuals or companies, for the whole nations and countries. Work in this direction can radically change the situation.".

In the opinion of the Director of the Institute, the ethical aspect of the Institute's products application is also important, as well as maintaining personal data security and privacy.

"When using such technologies, it is necessary that the rights of social networks users are not violated, personal data is not discredited," specified Denis Turdakov, head of the Information Systems Department ISP RAS. “We carry out scientific research, refine the "Texterra" and "Talisman" technologies. Comprehensive analysis of social networks is still hard, the breakthrough is yet to come, but it won't take long to wait."

At the moment, improvement of these software systems is continuing, their functionality and capacity are increasing. In the near future, the deployment of such products can dramatically change the face of the global information network and the culture of Internet communication.

All news