Information pressure – in the broadest sense, a consistent effect of information on the human psyche, in a narrow sense, public awareness with the aim of changing perception of the world and the events that take place, the adjustment or shift of the system of values.
Information pressure is a part of, on the one hand, the complex of instruments of external policy of foreign States on a par with economic incentives or sanctions – a kind of "political engineering" , on the other hand, the external ideological and value-oriented expansion and destructive information-psychological influence .
To strengthen our national security promotes the adoption of measures for the protection of Russian society from the pressure information, monitoring and evaluation of the information sector and the prevention of distribution of extremist content, violence, racial, religious and ethnic intolerance.
The relevance of evaluation information pressure is confirmed by the following:
- high intensity change of the information climate in the country and in separate regions,
- development of the telecommunications transmission systems (instant messengers, social networks, Internet of things),
- growing unfriendly influence on Russia in connection with the increasing importance of its role in the international arena,
- activities associated with the use of information and communication technologies for the dissemination and promotion of the ideology of fascism, extremism, terrorism and separatism.
High intensity change information, and therefore giving information to the target audience is also a major problem of assessment of information pressure which is determined by the following characteristics of information dissemination in online media and social networks:
- Social network content "sinks". Cannot find the message for a month or two ago;
- Time "active life" message is 4-6 hours. This is the time when the message is received by 98% of actions – comments, likes, etc.;
- Those who write himself, or at least adds comments to the message – 5%, others only borrow.
The next problem is a large amount of updated data. As the statistics shows media sources that distribute media content are effectively used video hosting. For example, at the beginning of 2016 video hosting service Youtube is visited daily by about 800 million people per month, 4 billion hours of video watched every month and this search engine was second in the Internet after Google Search. Every minute 72 hours of video being uploaded to the video hosting, a third of which contains a prohibited material (violence and propaganda).
These problems generate a number of tasks of collecting, recording, systematization, analysis, processing and storage of these data.
The problem of evaluation informational pressure
The preparatory stage of assessment information pressure is the process of determining the sources. Depending on the characteristics of the population and development and the quality of local information sources selects relevant sources (online media and social media users) in different categories of the target audience living in a separate area or region of the Russian Federation (on the order of 10-100 for each category). Using content analysis of the page revealed the relationship between sources and identifies the thematic group of sources, links to external sources and groups of sources that allows you to expand the map of the information sphere of the target audience (population of Russia).
The process of evaluation informational pressure is held with the aim of continuous monitoring and operational decision-making on carrying out stabilizing and counter measures for information security of the Russian population and includes two types of problems: problems of the operative information collection and problems of data processing.
The problem of efficient collection of information
The process of gathering information includes obtaining and structuring primary data. Going texts messages, relationships between users, links to external resources, photographic data, which are indexed and stored in the data store. Features of this phase are largely determined by the heterogeneity of the data used and mode of their treatment. System supporting the assessment in real-time, difficult to implement, and require high performance of the collection algorithm than using retrospective data collection. The reason is that most of the content is duplicated and, for comparison need to search all over the Internet and not in the indexed data store. The main problem of limiting accesses to information sources, an example of which is presented in table 1.
Table 1. Restriction of access to sources of information
Problems of data processing
System, seeking to make possible a systematic analysis of large data sets must take these issues into account, or at least not contradict them.
First, a large part of the work done in analysis is preprocessing of data. The data are messy, and cleaning, editing, merging, mixing and many other actions needed to benefit from. In particular, large data sets due to their inaccessibility for direct analysis people may require computational methods, just to find out what steps are pre-processing is required.
Even when it comes to performance optimization model for a typical data pipeline will need to spend a lot more time on design and feature selection than the selection and writing of algorithms. For example, when constructing a model for evaluating information pressure will have to choose from a wide variety of potential symptoms, such as: any fields you need to fill out the user information about the IP, date and time you access the site, logs keystrokes during the user's navigation on the website. They all have complexities that arise when converting to vectors suitable for machine learning algorithms. The system will need support for more flexible transformations than a simple transformation of the two-dimensional array of double-precision numbers in a mathematical model.
Secondly, the basic concept of the data processing is iteration. Modeling and analysis typically require multiple passes over the same data. One reason for this lies in the very nature of machine learning algorithms and statistical procedures. Popular optimization procedures like stochastic gradient descent or the method of maximization of the mathematical expectation, include repeated passes over the input data to achieve convergence. Iteration is also important when we talk about the sequence of data scientists.
Choosing the right signs, selection of the right algorithms, the implementation of proper checks on criteria of significance and finding the right hyperparameters – all this requires experimentation. The framework, which is necessary each time to read the same dataset from disk every time you open it, leads to delay and slow down the evaluation process and restricts the number of things you can try.
Third, the task does not end with the construction of a well-functioning model. Mechanisms of recommendations to address and systems for detecting pressure information in real time reaches the highest point in the information applications. They become part of a service, operating in the mode of operation, and require periodic (sometimes even in real time) adjustment.
Modern military-political situation, new threats to national security, including information pressure and also the modern features of the distribution of the information require a new technological approach to evaluation. Presents an overview of the main problems of the evaluation of the pressure information will enable a shift to effective prototyping of the technical system, and the constraints will improve its performance.
- The decree of the President of the Russian Federation from 31.12.2015 N 683 "On the strategy of national security of the Russian Federation".
- The speech of the Secretary of Security Council of the Russian Federation Igor Ivanov at the Assembly of the Council on foreign and defense policy on March 18, 2007
- Volkogonov D. A. Psychological war (Subversive actions of imperialism in the public consciousness). - M.: Voenizdat, 1984. - 320 p.
- Manoilo A.V. State information policy in conditions of information-psychological conflicts of high intensity and social risk Training manual. Moscow: MIFI, 2004. – 392 p.
- Zhuravlev A. L. Psychological impact: Mechanisms, strategies, capacity for resilience, ed. by A. L. Zhuravlev, N. D. Pavlova. - M.: Publishing house "Institute of psychology RAS", 2012. - 368 p. (Proceedings of the Institute of psychology Russian Academy of Sciences) ISBN 978-5-9270-0220-7
- 17-02-2017Planet Labs launched into space hundreds of tiny satellites around the clock to remove the entire surface of the Earth. Why is it important?
- 04-12-2016Nuclear flying monster. As in the Soviet Union was building nuclear aircraft
- 16-10-2016What modern nuclear warhead. Troubling questions about nuclear war
- 11-07-2016National Interest: the U.S. Navy is faced with challenges unprecedented since Soviet times
- 12-09-2010Many experts believe the best tank Merkava main battle tank in the world