Software & Systems

Articles of journal № 3 at 2013 year.

Order result by: Public date | Title | Authors |

1. Combining compile-time and run-time instrumentation for testing tools [№3 за 2013 год]
Authors: Iskhodzhanov T., Kleckner R., Stepanov E.
Visitors: 15720
Dynamic program analysis and testing tools typically require inserting extra instrumentation code into the program to test. The inserted instrumentation then gathers data about the program execution and hands it off to the analysis algorithm. Various analysis algorithms can be used to perform CPU profiling, processor cache simulation, memory error detection, data race detection, etc. Usually the instrumentation is done either at run time or atcompile time – called dynamic instrumentation and compiler instrumentation, respectively. However, each of these methods has to make a compromise between performance and versatil-ity when used in industry software development. This paper presents a combined approach to instrumentationwhich takes the best of the two worlds – the low run-time overhead and unique features of compile-time instrumentation and the flexibility of dynamic instrumentation. Wepresent modifications of two testing tools that benefit from thisapproach: AddressSanitizer and MemorySanitizer. We propose benchmarks to compare different instrumentation frameworks in conditions specific to hybrid instrumenta-tion. We discuss the changes we made to one of the state-of-the-art instrumentation frameworks to significantly improve the performance of hybrid tools.

2. Solution of a software structural construction problem for intelligent humidity sensor [№3 за 2013 год]
Author: Артемова С.В.
Visitors: 8186
The article considers the approach to solution of softwarestructural construction problem using the intellectual humidity sensor (IHS), which is an information-measuring system (IMS).When designing software and data support, the main modules of HIS are: an intellectual interface, a supervisor, survey sensors modules and measurement procedures mod-ules. The functioning of the IHS is based on using databases and knowledge bases. The IHS implementation requires a num-ber of algorithms, which should be placed in the modules to provide the designed IMS with specified properties. Such tasks of a choice belong to the class of NP-complete that can be solved only approximately. Therefore, for practical solving of the software structural construction problem it is offered to use heuristic methods, in particular, genetic algorithms. Method to solve the tasks includes following stages: forming the vector-line of algorithms distribution genes on the modules,description of a matrix of relations, interaction of algorithms using information elements, creating the matrix of alternatives ofalgorithm belonging to module, determining the target function. This approach can be used for designing data support and software for information-measuring or information management systems, that involves a large number of modules and algorithms.

3. Solving the problem of premises, buildings and constructions classification according to explosion and fire hazard categories using CAD [№3 за 2013 год]
Authors: Баранников Н.И., Сергеева М.А.
Visitors: 14473
The explosion and fire hazard object categories are the factor defining the requirements to designing and set-ting out a building. Correct definition of this category for buildings or constructions allows making optimum design decisions and preventing emergency situations with minimum necessary resource sosts. The article investigates the problem of categorization of premises, buildings and constructions according to explosion and fire hazard. The structure of the automated subsystem which is part of fire safety CAD is offered. The authors also offer the information support contents for a subsystem. The calculationalgorithms of explosion and fire hazard criteria using the rules of fire load optimization in CAD are considered.

4. Social networks analysis models and methods [№3 за 2013 год]
Author: Batura T.V.
Visitors: 24619
The paper represents the survey of the social network analysis methods. That analysis is used to study the in-teractions between network members, prediction their behavior, classification of the members, modeling information flow in networks. There are four main research areas: structural, resource, regulatory and dynamic. To solve the problems in social network analysis the following methods are used: graph and stochastic models, models of network evolution, methods in-volving ontologies, structural and relational models, machine learning methods, network visualization techniques, etc. This paper provides a brief description of popular computer socialnetworks, various approaches and models used in the network characteristics calculation, and lists the software applications that help automate the process of analysis. Several possible di-rections of further studies in this area are outlined: creation of social networks integrated theory, adaptation ofnatural lan-guage text processing methods to the online content, etc. In order to generalize the behavior of specific members ofthe net-work to the entire network, it is necessary to study methods for the detection and characterization of networks, patterns of distribution of these characteristics, the creation of methods to determine the cause of the participants’ interaction.

5. Software development risks management [№3 за 2013 год]
Authors: Бахтизин В.В., Кузиков А.А.
Visitors: 5245
The article examines the obstacles that are usual for software development teams working with the Scrum methodology and the impact on software quality. The authors propose a method aimed at risks minimizing when risks are connected with ineffective productivity time of Scrum-team members during Scrum iteration. The paper suggests the meas-ures that help in the risks quantitative assessment for late performances when the risks are linked with team members’ idle time and non-optimal decomposition of scheduled work.

6. Infrastructural model of professional Internet (Special Web) [№3 за 2013 год]
Authors: Беда И.А., Стюгин М.А., Носов И.А.
Visitors: 7155
The paper presents the website organization model that automatically ranks related information. The authors introduce an information and user quality index, which allows to efficiently search for relevant information and system users. The paper gives source data to develop algorithms for such anetwork self-organizing. One of the current major problems of today's Internet communication is that there is no possibility to carry out the semantic information and people search on pro-fessional expertise basis. This problem solution can provide an impetus to develop professional communities and communi-cations on the Internet. This problem is expected to be solved by creating professional Internet infrastructure that includes such elements as specialization tree; impact factors according to the specialization, web page and the user; user activity index registered in information field that indicates movements on web pages; positive and negative reactions to the information. An important science component in infrastructure construction is design when context indicators of web resources and users are corrected automatically, without outside interference. This process requires a system dynamics algorithm that is formed by using heuristic algorithms with input data obtained based on expert assessments. To achieve dynamics in a system, it is nec-essary to form the primary impact factors of the system. Questions to test the competence open to evaluation and comparison between other users in the system can solve this problem, if you compare them with specialization tree. This work is current-ly performed by the staff of the Siberian Federal University together with the company LLC Kairos.

7. Program to solve the optimal ball reorientation problem [№3 за 2013 год]
Author: Бесчастный И.Ю.
Visitors: 7701
The article considers the optimal problem of a rolling ball on a plane with twisting but without slipping. It is a modification of a classical model problem from geometric mechanics, when it is required to optimally transfer the ballfrom a given initial state to a terminal one without twistingor slipping. Nowadays both problems are open and even in the non-twisting case the optimal trajectories in the reorientation sub-problem are unknown (when it is required to transfer the ball to the initial point with a new orientation). But if twisting is allowed, the problem becomes significantly simplier. The paper considers the program interface for optimal reorientation in the twisting case. Using some necessary optimality conditions obtained in [10] we construct a solution algorithm, whichwas realized as a program interface in Wolfram Mathematica.The algorithm considered in this paper can be used for motion planing of a spherical robot.

8. Approximate valuation of potentially dangerous objects security. Object security structural parameters [№3 за 2013 год]
Authors: Borovsky A.S., Тарасов А.Д.
Visitors: 13304
The article considers the decision-making problem for potentially dangerous objects physical protection sys-tem structure. It defines the stage of a problem using experts’ knowledge – conceptual designing, that includes object security analysis, practical recommendations for physical protection system creation, options of engineering security facilities struc-ture. The problem of physical protection system synthesis is formulated. There are examples of various problem decision methods, the problem of objects security performance evaluation by engineering security facilities. The problem includes ex-perts knowledge discrepancy and ambiguity in data finding for object security analysis. The purpose of the article is working out models and methods for objects security analysis. The method based on representation the object to be protected in graph form. The general properties of object’s graph are considered.There is a description of a structural protection degree. The degree allows indicating efficiency level for security facilities located on the intruder path, when there is an attempt of infil-tration on object and moving on object’s territory for thecriminal purpose. The authors analyze a chance for an intruder to reach one of object’s critical elements without detectingby security facilities and without stopping by physicalbarriers. The structural defense degree shows a chance of intruder detection and stopping on the most vulnerable path to a critical element. The paper estimates the possibility of using Dijkstra's algorithm to search the least protected path, and develops themodified algorithm for structural dsecurity degree searching. The methods of expert information processing in fuzzy numbers form are described. The example of searching structural protectiondegree using the program is shown.

9. The homogeneity analysis application to visualize and analyse bibliometrics and scientometrics data [№3 за 2013 год]
Authors: Бурилин А.В., Гордеев Р.Н.
Visitors: 10431
The ensembles of trees, especially random forest have proved themselves to produce accurate predictions for solving regression and classification tasks. Perhaps unjustified, the main criticism of this approach is that these methods act as black boxes and do not provide additional information on the objects that are not participating in the training set. This paper considers a multi-class classification and shows that the homogeneity analysis used mainly in psychometrics can be used to provide effective visualization of the ensemble of trees, including visualization of the new observations that were not included in the training set. Observations and rules (nodes) of the ensemble of trees are placed in a bipartite graph connecting each observation with all the rules (nodes) that satisfies it. The location of the graph in this case is chosen according to minimizing the sum of edges lengths squares with certain limitations.

10. Partitioning method and its application when classifying heterogeneous information [№3 за 2013 год]
Authors: Бурилин А.В., Гордеев Р.Н., Борисов П.А.
Visitors: 10083
In our previous work we considered the analysis of homogeneity and believe that it has great potential for ef-fective visualization of ensemble of trees and similar machine learning algorithms. However, there are at least two drawbacks to this approach: the computational problems may arise if the number of training observations is very large and, more im-portantly, the accuracy of prediction in two-dimensional embeddings often much worse than in the original ensemble oftrees algorithms, this means that significant information is lostin low-dimensional embeddings. We present a simple extension analysis of homogeneity called sectioning, which often solves the above mentioned prob-lems in the case of multi-class ranking and can lead to a significant improvement in prediction accuracy.

| 1 | 2 | 3 | 4 | 5 | 6 | Next → ►