卷 21, 编号 6 (2022)
Articles
SPC RAS — 45 Years of Scientific Activity
摘要
45 years ago, had started the history of the St. Petersburg Institute for Informatics and Automation of the Russian Academy of Sciences (SPIIRAS), when by the Order of the Council of Ministers of the USSR dated December 19, 1977 and by the Decree of the Presidium of the Academy of Sciences of the USSR dated January 19, 1978, the Leningrad Research Computer Center of the USSR Academy of Sciences (LRCC) had been established, that later in 1985 was transformed by the decision of the Presidium of the USSR Academy of Sciences into the Leningrad Institute for Informatics and Automation of the USSR Academy of Sciences (LIIAS), then in 1991, as stipulated by the return to the city of its historical name, the Institute acquired its current name of SPIIRAS. By its 45th anniversary, the Institute for Informatics (LRCC, LIIAS, SPIIRAS), having united with five leading academic entities of the North-West of Russia, in 2020 turned into the St. Petersburg Federal Research Center of the Russian Academy of Sciences (SPC RAS). With no exaggeration, a distinctive mark of the SPC RAS scientific team is systemic interdisciplinary research. The fundamental results received by the SPC RAS scientists in computer science, cybernetics, artificial intelligence, robotics, security, ecology, agriculture and innovation and investment development of territories fit well in forming the applied solutions that regard the digital transformation of agroecological production, strengthening food, environmental and information security of citizens of our country.
Informatics and Automation. 2022;21(6):1085-1096
1085-1096
Artificial intelligence, knowledge and data engineering
Analysis of infoware and software for human affective states recognition
摘要
The article presents an analytical review of research in the affective computing field. This research direction is a component of artificial intelligence, and it studies methods, algorithms and systems for analyzing human affective states during interactions with other people, computer systems or robots. In the field of data mining, the definition of affect means the manifestation of psychological reactions to an exciting event, which can occur both in the short and long term, and also have different intensity. The affects in this field are divided into 4 types: affective emotions, basic emotions, sentiment and affective disorders. The manifestation of affective states is reflected in verbal data and non-verbal characteristics of behavior: acoustic and linguistic characteristics of speech, facial expressions, gestures and postures of a person. The review provides a comparative analysis of the existing infoware for automatic recognition of a person’s affective states on the example of emotions, sentiment, aggression and depression. The few Russian-language, affective databases are still significantly inferior in volume and quality compared to electronic resources in other world languages. Thus, there is a need to consider a wide range of additional approaches, methods and algorithms used in a limited amount of training and testing data, and set the task of developing new approaches to data augmentation, transferring model learning and adapting foreign-language resources. The article describes the methods of analyzing unimodal visual, acoustic and linguistic information, as well as multimodal approaches for the affective states recognition. A multimodal approach to the automatic affective states analysis makes it possible to increase the accuracy of recognition of the phenomena compared to single-modal solutions. The review notes the trend of modern research that neural network methods are gradually replacing classical deterministic methods through better quality of state recognition and fast processing of large amount of data. The article discusses the methods for affective states analysis. The advantage of multitasking hierarchical approaches is the ability to extract new types of knowledge, including the influence, correlation and interaction of several affective states on each other, which potentially leads to improved recognition quality. The potential requirements for the developed systems for affective states analysis and the main directions of further research are given.
Informatics and Automation. 2022;21(6):1097-1144
1097-1144
Recurrent Neural Networks with Continuous Learning in Problems of News Streams Multifunctional Processing
摘要
The main task of using neural networks is the prompt and accurate solution of various creative tasks, including the analysis and synthesis of news flows, while maintaining the continuity of learning. The result of such processing can be digests, filtered news streams, as well as event forecasts that allow for proactivity in management decisions. Known methods of news processing by neural networks and technical solutions that implement them do not fully provide a solution to the problems that arise in this area. It is necessary to expand their functionality, and improve the space-time signal binding in recurrent neural networks. When processing news flows, simultaneously with continuous training of recurrent neural networks, selection, recognition, restoration, prediction and synthesis of news should be carried out. To reduce the severity of the problem, a promising method of multifunctional processing of news flows is proposed using recurrent neural networks with a logical organization of layers and continuous learning. The method is based on the development of associative processing of textual information in streaming recurrent neural networks with controlled elements. The key features of this method are the multifunctional processing of information flows with changing laws of news appearance. The method provides for operational selection, recognition, restoration, forecasting and synthesis of news based on deep associative continuous processing of links between text elements. The neural network system that implements the proposed method differs from the known solutions by new elements, connections between them, as well as by the functions performed. The results of the experiments confirmed the extended functionality of the method. New features of processing news texts by streaming RNNs are revealed. The proposed solutions can be used to create a new generation of intelligent systems not only for word processing, but also for other types of information.
Informatics and Automation. 2022;21(6):1145-1168
1145-1168
Method and Models of Extraction of Knowledge from Medical Documents
摘要
The paper analyzes the problem of extracting knowledge from clinical recommendations presented in the form of semi-structured corpora of text documents in natural language, taking into account their periodic updating. The considered methods of intellectual analysis of the accumulated arrays of medical data make it possible to automate a number of tasks aimed at improving the quality of medical care due to significant decision support in the treatment process. A brief review of well-known publications has been made, highlighting approaches to automating the construction of ontologies and knowledge graphs in the problems of semantic modeling of a problem-oriented text corpus. The structural and functional organization of the system of knowledge extraction and automatic construction of an ontology and a knowledge graph of a problem-oriented corpus for a specific subject area is presented. The main stages of knowledge extraction and dynamic updating of the knowledge graph are considered: named entity extraction, semantic annotation, term and keyword extraction, topic modeling, topic identification, and relationship extraction. The formalized representation of texts was obtained using a pre-trained BERT transformer model. The automatic selection of triplets "object" - "action" - "subject" based on part-of-speech markup of the text corpus was used to construct fragments of the knowledge graph. An experiment was carried out on a corpus of medical texts on a given topic (162 documents of depersonalized case histories of patients of a pediatric center) without preliminary markup in order to test the proposed solution for extracting triplets and constructing a knowledge graph based on them. An analysis of the experimental results confirms the need for a deeper markup of the corpus of text documents to take into account the specifics of medical text documents. For an unmarked corpus of texts, the proposed solution demonstrates satisfactory performance in view of the selection of atomic fragments included in the automatically generated ontology.
Informatics and Automation. 2022;21(6):1169-1210
1169-1210
Algorithms and Measuring Complex for Classification of Seismic Signal Sources, Determination of Distance and Azimuth to the Point of Excitation of Surface Waves
摘要
Machine learning and digital signal processing methods are used in various industries, including in the analysis and classification of seismic signals from surface sources. The developed wave type analysis algorithm makes it possible to automatically identify and, accordingly, separate incoming seismic waves based on their characteristics. To distinguish the types of waves, a seismic measuring complex is used that determines the characteristics of the boundary waves of surface sources using special molecular electronic sensors of angular and linear oscillations. The results of the algorithm for processing data obtained by the method of seismic observations using spectral analysis based on the Morlet wavelet are presented. The paper also describes an algorithm for classifying signal sources, determining the distance and azimuth to the point of excitation of surface waves, considers the use of statistical characteristics and MFCC (Mel-frequency cepstral coefficients) parameters, as well as their joint application. At the same time, the following were used as statistical characteristics of the signal: variance, kurtosis coefficient, entropy and average value, and gradient boosting was chosen as a machine learning method; a machine learning method based on gradient boosting using statistical and MFCC parameters was used as a method for determining the distance to the signal source. The training was conducted on test data based on the selected special parameters of signals from sources of seismic excitation of surface waves. From a practical point of view, new methods of seismic observations and analysis of boundary waves make it possible to solve the problem of ensuring a dense arrangement of sensors in hard-to-reach places, eliminate the lack of knowledge in algorithms for processing data from seismic sensors of angular movements, classify and systematize sources, improve prediction accuracy, implement algorithms for locating and tracking sources. The aim of the work was to create algorithms for processing seismic data for classifying signal sources, determining the distance and azimuth to the point of excitation of surface waves.
Informatics and Automation. 2022;21(6):1211-1239
1211-1239
Digital information telecommunication technologies
A Comparative Study by Simulation of OSPF and EIGRP Routing Protocols
摘要
Computer networks are based on technology that provides the technical infrastructure where routing protocols are used to transmit packets over the Internet. Routing protocols define how routers communicate with each other by distributing information. They are used to describe how routers communicate with each other, learn available routes, build routing tables, make routing decisions, and share information between neighbors. The main purpose of routing protocols is to determine the best route from source to destination. A particular case of a routing protocol operating within an autonomous system is called an internal routing protocol (IGP – Interior Gateway Protocol). The article analyzes the problem of correctly choosing a routing protocol. Open Shortest Path First (OSPF) and Enhanced Interior Gateway Routing Protocol (EIGRP) are considered leading routing protocols for real-time applications. For this they are chosen to be studied. The main objective of the study is to compare the proposed routing protocols and to evaluate them based on different performance indicators. This assessment is carried out theoretically – by analyzing their characteristics and action, and practically – through simulation experiments. After the study of the literature, the simulation scenarios and quantitative indicators by which the performance of the protocols is compared are defined. First, a network model with OSPF is designed and simulated using the OPNET Modeler simulator. Second, EIGRP is implemented in the same network scenario and a new simulation is done. The implementation of the scenarios shall collect the necessary results and analyze the operation of the two protocols. The data shall be derived and an assessment and conclusion shall be made against the defined quantitative indicators.
Informatics and Automation. 2022;21(6):1240-1264
1240-1264
Microservice Architecture of Virtual Training Complexes
摘要
The task of automating and reducing the complexity of the process of developing virtual training complexes is considered. The analysis of the subject area showed the need to move from a monolithic to a service-oriented version of the architecture. It is found that the use of a monolithic architecture in the implementation of virtual training complexes limits the possibility of modernizing the system, increases its software complexity, and makes it difficult to implement an interface for managing and monitoring the training process. The general concept of the microservice architecture of virtual training complexes is presented, and definitions of the main and secondary components are given. The scientific novelty of the research lies in the transition from the classical monolithic architecture in the subject area of the HTC to the microservice architecture; eliminating the shortcomings of this approach by implementing a single protocol for the exchange of information between modules; separation of network interaction procedures into software libraries to unify and improve the reliability of the system. The use of isolated, loosely coupled microservices allows developers to use the best technologies, platforms and frameworks for their implementation; separate the graphical interface of the simulator instructor from the visualization and virtual reality system; provide the ability to flexibly replace the main components (visualization, interface, interaction with virtual reality) without changing the architecture and affecting other modules. The decomposition of the structural model of the microservice architecture is carried out, and the specifics of the functioning of the main components are presented. The implementation of microservices networking libraries and a JSON-based data exchange protocol is considered. The practical significance of the proposed architecture lies in the possibility of parallelization and reducing the complexity of the development and modernization of training complexes. The features of the functioning of the systems implemented in the proposed microservice architecture are analyzed.
Informatics and Automation. 2022;21(6):1265-1289
1265-1289
Information security
Data Generation for Modeling Attacks on UAVs for the Purpose of Testing Intrusion Detection Systems
摘要
Today, issues related to ensuring the safety of UAVs are very relevant. Researchers need to develop new protection methods to detect attacks in a timely manner and implement mitigation measures. The authors propose a new concept of attack detection "from inside" the UAV. The idea is to analyze the cyber-physical parameters of the UAV, which may indicate an attack, and its possible consequences. It was determined that to detect an attack and determine the consequences to which it can lead, it is necessary to control not only the initial parameters, but also the internal cyber-physical parameters of the UAV. This will allow predicting the possible consequences of an attack and taking emergency measures. A scheme of the impact of an attack on UAVs and the relationship with security incidents, built using an ontological approach, has been worked out. Two main essences of the UAV are considered - the physical and digital aspects of the UAV. Examples of chains of attacks leading to various consequences are also shown. In the review part, the analysis of methods and algorithms for detecting spoofing attacks using data generators is carried out, based on which conclusions are drawn about their advantages and disadvantages. Further, based on the experiments performed, the authors propose a method for assessing the quality of data and a method for generating anomalous data sets similar to real attack data, which can be used to develop and test methods for detecting and blocking attacks. The architecture of the experimental stand, which was used in the framework of full-scale simulation, is described. At this stand, designed to parse GPS spoofing attacks (GPS spoofing), several scenarios of a normal flight, and then several attack scenarios, were tested. Based on the results of the experiments, a method has been proposed that allows simulating the data corresponding to the attack with the required accuracy. A method for assessing the quality of fake data has also been proposed.
Informatics and Automation. 2022;21(6):1290-1327
1290-1327
Anomaly and Cyber Attack Detection Technique Based on the Integration of Fractal Analysis and Machine Learning Methods
摘要
In modern data transmission networks, in order to constantly monitor network traffic and detect abnormal activity in it, as well as identify and classify cyber attacks, it is necessary to take into account a large number of factors and parameters, including possible network routes, data delay times, packet losses and new traffic properties that differ from normal. All this is an incentive to search for new methods and techniques for detecting cyber attacks and protecting data networks from them. The article discusses a technique for detecting anomalies and cyberattacks, designed for use in modern data networks, which is based on the integration of fractal analysis and machine learning methods. The technique is focused on real-time or near-real-time execution and includes several steps: (1) detecting anomalies in network traffic, (2) identifying cyber attacks in anomalies, and (3) classifying cyber attacks. The first stage is implemented using fractal analysis methods (evaluating the self-similarity of network traffic), the second and third stages are implemented using machine learning methods that use cells of recurrent neural networks with a long short-term memory. The issues of software implementation of the proposed technique are considered, including the formation of a data set containing network packets circulating in the data transmission network. The results of an experimental evaluation of the proposed technique, obtained using the generated data set, are presented. The results of the experiments showed a rather high efficiency of the proposed technique and the solutions developed for it, which allow early detection of both known and unknown cyber attacks.
Informatics and Automation. 2022;21(6):1328-1358
1328-1358
Robotics, automation and control systems
On Theoretical Foundations of Aerolimnology: Study of Fresh Water Bodies and Coastal Territories Using Air Robot Equipment
摘要
The integration of the methodological basis of several different sciences in interdisciplinary research is a characteristic feature of new mechanisms for solving modern applied problems. The emerging theoretical foundations of aerolimnology, as a new scientific direction, are considered from the point of view of the contribution of three key sciences to it: limnology, informatics and robotics. Classifications of methods and approaches of limnological research, airborne robotic means, and information technologies that are promising for solving problems in the field of aerolimnology are given. The task of the scientific direction of aerolimnology is formulated as the study of the possibilities and limitations of combined methods of remote sensory measurement, robotic sampling and analytical study of the parameters of freshwater ecosystems to monitor and predict the dynamics of their development. Among the main areas of aerolimnological research, the following are distinguished: the construction of orthophotomaps and photogrammetric spatial models of the bottom topography and individual elements of the bottom landscape and coastal zone of various scales; geological and geophysical mapping of the underwater part of the coastal zone; the study of phytoplankton, in particular, the "bloom" of water caused by cyanobacteria; study of distribution and migration of large representatives of hydrofauna; study of temperature fields and processes of redistribution of water masses. The limitations imposed on the use of unmanned aerial vehicles (UAVs) in sampling and monitoring coastal water areas are discussed, primarily weather-climatic, temporal, spatial, and technical. The advantage of using unmanned aerial vehicles in aerolimnology is justified by an increase in the speed of data acquisition, the possibility of approaching hard-to-reach and territorially remote objects, and a decrease in the influence of the human factor. The scientific novelty of the presented research consists in an attempt to integrate interdisciplinary knowledge when using unmanned aerial vehicles and processing the obtained data based on artificial intelligence technologies in the study of limnological objects and processes. The important role of geoinformation systems is noted and examples of maps of shore typification and geomorphology of Lake Ladoga are given, posted on the website of the Center for the Collective Use of Scientific Equipment "North-Western Center for Monitoring and Forecasting the Development of Territories" of the St. Petersburg Federal Research Center of the Russian Academy of Sciences. The main stages of the methodology for conducting aerolimnological studies using interdisciplinary approaches based on limnology, informatics and robotic tools operating in different environments are considered.
Informatics and Automation. 2022;21(6):1359-1393
1359-1393