The methodology of automatic detection of the event basis of information operations, reflected in thematic information flows, is described. The presented methodology is based on the technologies for identifying information operations, the formation of the terminological basis of the subject area, the application of cluster analysis with cluster centroids, determined by analyzing the terminology of the information flow. The clusters formed in this way reflect the main events occurring during the information operations and reveal the technique for their implementation.
At present, the Internet space becomes a battlefield, on which there are numerous information wars, individual information operations. Information operations are defined as "actions aimed at influencing information and information systems of the enemy and protecting their own information and information systems" [1]. Information operations are components and support for more general processes. At the same time, the arena of information operations is the information space, which, on the one hand, is the place of information battles, and on the other hand, the environment for displaying real combat operations [2]. In this case, information operations in practice are supported by numerous events, processes, actions (under the events within the framework of this work we will understand a significant incident, phenomenon or manifestation of other activity as a fact of public or private life). Analysis of the reflection of events in the Internet space makes it possible to identify the participants in information confrontations, methods of information impact, to uncover the technique of implementing information operations.
The purpose of this work is to create and justify a methodology for identifying the event basis of information operations. The implementation of this methodology will let us determine the time frame of the information operation, identify the main events that accompany the information operation and see the techniques of information impact.
When studying information operations it is necessary to determine objective criteria, and as one such, one can consider the dynamics of the distribution of information plots in the corresponding fragment of the information space. Numerous scientific works [3], [4], [5], [6] have been devoted to the investigation of the dynamics of information flows, it is shown that in typical situations the dynamics of the distribution of news, information plot is characterized by the nature of a “burst”, waves with an obvious period of increasing its influence and smooth decrease. At the same time, the question of determining the event basis for information operations remains open.
It is assumed that a systematic violation of the typical dynamics of some thematic information flows in the open information space may indicate information operations [7]. In the study of information operations, much attention is also paid to the analysis of the dynamics of information flows, in [8] a typical template of the information operation is presented (Figure 1), which allows using available analytical tools, for example, correlation analysis [9].
To get the dynamics of the thematic flow on a certain topic, you can use content monitoring systems. As the system of content monitoring, the authors selected the InfoStream system, which currently covers 10,000 sources of information in Russian, Ukrainian and English. The database of the system receives more than 100 thousand documents daily. The InfoStream system provides a search, as well as viewing the list and full texts of relevant documents.
In the example shown in Fig. 2 a fragment of the system interface through which the request for the referendum on the UK exit from the European Union (abbreviated Brexit from the combination of the words Britain -Britain and English Exit -exit) processed during 2016 (the period of the study June-July 2016) is shown. As a result, a thematic information array was created, covering 43697 documents. In Fig. 3 a graph of the dynamics of this information flow, as well as the result of its smoothing with a window in 7 days is shown.
To determine the degree of “proximity” of the dynamics of the thematic information flow to the information operation, the method proposed in [9] was used. The idea is to compare parts of a series of information flow dynamics with a certain pattern on different scales. For this, the correlation between the time series part and some template (the scaled part of the graph shown in Figure 1) is calculated:
xx . That is, the parameter k corresponds to the pattern shift, and the parameter k corresponds to the number of points in the pattern and in the considered segment of the row. The parameter k in this case is analogous to the scale. In this case, for calculation , C l k the k points of the series and the length pattern are used. When visualizing, the most light colors correspond to the highest values. Fig. 4 shows a Correlogram of the time series corresponding to the dynamics of the information flow on the Brexit topic and the template shown in Fig. 1.
The analysis of the given correlogram makes it possible to narrow the time frame in the study of the thematic information flow.
To determine the event basis of information operations, reference words are extracted from the information flow documents, for which several algorithms can be used [10]. The authors, in particular, used the TF-IDF algorithm implemented in the InfoStream system. Then, the significant information
This content is AI-processed based on open access ArXiv data.
