FEASIBILITY OF USING METHODS OF ALMOST PERIODIC FUNCTIONS, WAVELET ANALYSIS AND HURST SELF-SIMILARITY FOR PREDICTING NEWS EVENTS IN THE INFORMATION SPACE
Journal Title: Современные информационные технологии и ИТ-образование - Year 2017, Vol 13, Issue 1
Abstract
In the present paper we consider the feasibility of using methods of almost periodic functions, wavelet analysis and Hurst self-similarity to analyze in length of time the behavior spectra of vectors defining the position of the news reports clusters in the information space. The essence of the authors approach is to apply the methods of mathematical linguistics (text markup, normalization, comment) to create a dictionary and a collection of news text messages tied to the time scale. This makes it possible to create a vector representation for each newsletter, using standard methods. It is proposed to introduce the concept of a director (a conditional axis characterizing the basic direction of all vectors) for the entire set of vectors presented in the article. Time progress in metrics (cosine of the angle) of the vectors defining the position of the clusters centers relative to directors form spectra of information processes. The analysis by the methods of almost - periodic functions, wavelet analysis and Hurst self-similarity can help identify the presence of recurrence of certain social events groups, and thus predict their possible manifestation in the future.
Authors and Affiliations
Dmitriy Zhukov, Olga Novikova, Anton Alyoshkin
Keywords
Related Articles
- EP ID EP320789
- DOI 10.25559/SITITO.2017.1.445
- Views 119
- Downloads 0
COGNITIVE GOALS SIMULATION BY FORMALIZED KNOWLEDGE PROCESSING OPERATIONS
Searching the foundations for the formal concepts of intelligent systems creation and application includes development of the unified tools that integrate experience of the implementation of application-oriented knowledg...
PARTICULAR QUALITIES OF THE DEVELOPMENT AND APPLICATION OF FDM-TECHNOLOGY FOR CREATING AND PROTOTYPING 3D-OBJECTS
The article gives comparative analysis of additive technologies used in prototyping 3D-objects, and features of the FDM-technology are characterized. The parameters of the 3D printer using FDM-technology are described, a...
EXPERIENCE IN DEVELOPMENT, TRENDS IN THE DEVELOPMENT AND IMPLEMENTATION OF INFORMATION SYSTEMS SUPPORTING THE MAIN EDUCATIONAL PROCESS
This article contents experience of use and description of tendencies of educational information systems extension with emphasis on institutions of higher education. In the context of the emergence of new educational nee...
ON APPROACHES TO ANALYZING DEMOGRAPHIC DATA USING MACHINE LEARNING
Demographic data are fairly accessible data sets that can be used for analysis with the use of modern technologies of artificial intelligence and machine learning (ML). However, they cannot be used for these purposes wit...
THE BASIC MODES OF THE INTRUSION PREVENTION SYSTEM (IDS/IPS SURICATA) FOR THE COMPUTING CLUSTER
This article aims to test the Suricata intrusion prevention system and consider its use as a primary or additional tool for securing the computing cluster. For solve this task, we tested the operation of this system in t...