CONTENTS
-
-
-
-
-
A. I. Goglachev
South Ural State University (National Research University, 454080, Chelyabinsk, Russia
DISCOVERY IN TIME SERIES BASED ON THE CONCEPT OF CANONICAL FEATURES
DOI: 10.24412/2073-0667-2026-2-5-31
EDX: ZMIIHW
This article presents a new parallel algorithm for time series summarization patterns discovery, PDSS (Parallel Discovery of Semantic Snippets), based on the concept of canonical time-series characteristics (CATCH22). Summarization patterns are a set of subsequences (continuous intervals) of a time series that reflect the activities of the object under study and enable time series labeling. The developed algorithm for summary patterns discovery involves four stages: calculating feature profiles of time series subsequences, normalizing features, calculating a distance matrix, and searching for patterns. PDSS enables the search for summary patterns and labeling of time series from various subject areas, including multivariate time series. Computational experiments on two standard datasets demonstrate that the proposed algorithm outperforms state-of-the-art analogs in terms of performance and accuracy, based on the FI score, by an order of magnitude and 33 %, respectively.
Key words: time series, pattern subsequence mining, GPU, canonical features.
References
1. Volkov I., Radchenko G., Tchcrnykh A. Digital Twins, Internet of Things and Mobile Medicine: A Review of Current Platforms to Support Smart Healthcare /7 Programming and Computer Software
2021. Vol. 47, N 8. P. 578 590. DOI: 10.1134/S0361768821080284.
2. Liu Y. Scalable Multivariate Time-Series Models for Climate Informatics /7 Computing in Science V Engineering. 2015. Vol. 17, N 6. P. 19 26. DOI: 10.1109/MCSE.2015.126.
3. Arruda H. М., Bavaresco R. S., Kunst R. et.al. Data Science Methods and Tools for Industry 4.0: A Systematic Literature Review and Taxonomy //Sensors. 2023. Vol. 23, N 11. P. 5010. DOI: 10.3390/S23115010.
4. Kumar S., Tiwari P., Zymbler M. Internet of Things is a revolutionary approach for future technology enhancement: a review /7 .Journal of Big Data. 2019. Vol. 6, N 1. DOI: 10.1186/740537¬
019-0268-2.
5. Imani S., Madrid F., Ding W., Croutcr S., and Keogh E. Matrix Profile XIII: Time Series Snippets: A New Primitive for Time Series Data Mining /7 2018 IEEE International Conference on Big Knowledge, ICBK 2018. Nov. 17 18, 2018, Singapore. P. 382 389. DOI: 10.1109/ICBK.2018.00058.
6. Gharghabi S., Imani S., Bagnall A., Darvishzadeh A., and Keogh E. An ultra-fast time series distance measure to allow data mining in more complex real-world deployments /7 Data Mining and Knowledge Discovery. 2020. Vol. 34, N 4. P. 1104 1135. DOI: 10.1007/sl0618-020-00695-8.
7. Lubba С. H., Sethi S. S., Knautc P. et. al. catch22: CAnonieal Time-series CHaracteristics: Selected through highly comparative time-series analysis /7 Data Mining and Knowledge Discovery.
8. Vol. 33, N 6. P. 1821 1852. DOI: 10.1007/sl0618-019-00647-x.Mueen A., Keogh Е., Zhu Q., Cash S., and Westover В. Exact Discovery of Time Series Motifs// Proceedings of the 2009 SIAM International Conference on Data Mining. Society for Industrial and Applied Mathematics. SDM 2009. Apr. 30-May 2, 2009, Sparks, Nevada, USA. DOI: 10.1137/1.9781611972795.41.
9. Zvmbler M. L., Kraeva Y. A. Discovery of Time Series Motifs on Intel Manv-Core Systems // Lobachevskii Journal of Mathematics. 2019. Vol. 40, N 12. P. 2124-2132. DOI: 10.1134/S199508021912014X.
10. Narang A., Bhattacherjee S. Parallel Exact Time Series Motif Discovery // Euro-Par 2010 - Parallel Processing. Springer Berlin Heidelberg, 2010. Aug. 31-Sep. 3, 2010, Ischia, Italy. P. 304-315. DOI: 10.1007/978-3-642-15291-7^28.
11. Zvmbler M.L., Kraeva Y. A. Parallel algorithm for time series motif discovery on graphic processor // Bulletin of the South Ural State University. Series: Computational Mathematics and Software Engineering. 2020. T. 9, N 3. DOI: 10.14529/cmse200302.
12. Ye L., Keogh E. Time series shapelets: a new primitive for data mining // Proceedings of the 15th ACM SIGKDD international conference on Knowledge discovery and data mining. KDD09. June 28-July 1, 2009, Paris, Prance. P. 947-956. DOI: 10.1145/1557019.1557122.
13. Indvk P., Koudas N., Muthukrishnan S. Identifying Representative Trends in Massive Time Series Data Sets Using Sketches // Proceedings of the 26th International Conference on Very Large Data Bases. VLDB ’00. Sep. 10-14, Cairo, Egypt. P. 363-372.
14. Ermshaus A., Schafer P., Leser U. ClaSP: parameter-free time series segmentation // Data Mining and Knowledge Discovery. 2023. Vol. 37, N 3. P. 1262-1300. DOI: 10.1007/sl0618-023-00923-x.
15. Gharghabi S., Yeh C., Ding Y. et.al. Domain agnostic online semantic segmentation for multi-dimensional time series // Data Mining and Knowledge Discovery. 2018. Vol. 33, N 1. P. 96-130. DOI: 10.1007/sl0618-018-0589-3.
16. Yeh C., Zhu Y., Ulanova L. et. al. Matrix Profile I: All Pairs Similarity Joins for Time Series: A Unifying View That Includes Motifs, Discords and Shapelets // 2016 IEEE 16th International Conference on Data Mining. ICDM 2016. Dec. 12-15, 2016, Barcelona, Spain. P. 1317-1322. DOI: 10.1109/ICDM.2016.0179.
17. Jin X., Han J. К-Means Clustering // Encyclopedia of Machine Learning. Springer US, 2011. P. 563-564. DOI: 10.1007/978-0-387-30164-8^425.
18. Truong C., Oudre L., Vavatis N. Selective review of offline change point 1 detection methods // Signal Processing. 2020. Vol. 167. P. 107299. DOI: 10.1016/j.sigpro.2019.107299.
19. Zhou W., Chan Y. E., Foo C. S. et. al. High-Resolution Digital Phenotypes From Consumer Wearables and Their Applications in Machine Learning of Cardiometabolic Risk Markers: Cohort Study // Journal of Medical Internet Research. 2022. Vol. 24, N 7. P. e34669. DOI: 10.2196/34669.
20. Diraco G., Siciliano P., Leone A. Behavioral Change Prediction from Physiological Signals Using Deep Learned Features // Sensors. 2022. Vol. 22, N 9. P. 3468. DOI: 10.3390/s22093468.
21. Gamberini A., Abdallah I. Active trailing edge flap system fault detection via machine learning // Wind Energy Science. 2024. Vol. 9, N 1. P. 181-201. DOI: 10.5194/wes-9-181-2024.
22. Tafazoli S., Lu Y., Wu R., Srinivas T. et. al. C22MP: the marriage of catch22 and the matrix profile creates a fast, efficient and interpretable anomaly detector //
https://doi.org/10.1007/ s 10115-024-02107-5 Knowledge and Information Systems. 2024. Vol. 66, N 8. P. 4789-4823.
23. N. Satish, M. Harris, and M. Garland. Designing efficient sorting algorithms for manvcore GPUs // Proceedings of the 2009 IEEE International Symposium on Parallel k, Distributed Processing. IEEE Comput. Soc. Press. IPDPS ’09. May' 23-29, Rome, Italy. P. 1-10. DOI: 10.1109/IPDPS.2009.5161005.
24. Wang X., Wirth A., Wang L. Structure-Based Statistical Features and Multivariate Time Series Clustering // Seventh IEEE International Conference on Data Mining. ICDM 2007. Oct. 28-31, 2007, Omaha, Nebraska, USA. P. 351-360. DOI: 10.1109/ICDM.2007.103
26. Bilenko R. V., Dolganina N. Yu., Ivanova E. V., Rekachinskv A. I. High-performance Computing Resources of South Ural State University// Bulletin of the South Ural State University. Series: Computational Mathematics and Software Engineering. 2022. T. 11, N 1. DOI: 10.14529/cmse220102.
27. Zvmbler М., Goglachev A. Fast Summarization of Long Time Series with Graphics Processor // Mathematics. 2022. Vol. 10, N 10. P. 1781. DOI: 10.3390/mathl0101781.
28. Reiss A., Strieker D. Introducing a New Benchmarked Dataset for Activity Monitoring // 2012 16th International Symposium on Wearable Computers. ISWC 2012. June 18-22, 2012. DOI: 10.1109/ISWC.2012.13.
29. Harris M. Optimizing parallel reduction in CUDA // NVIDIA Developer Technol. 2007. Vol. 2, N 4. P.70
Bibliographic reference: Goglachev A.I. A Parallel Algorithm for Summarization Pattern Discovery in Time Series Based on the Concept of Canonical Features//"Problems of informatics", 2026, № 2, pp..5-31. DOI: 10.24412/2073-0667-2026-2-5-31.
A. A. Matsegora*, D.A. Migov, A. S. Rodionov
Institute of Computational Mathematics and Mathematical Geophysics SB RAS, 630090, Novosibirsk, Russia
*Novosibirsk State University, 630090, Novosibirsk, Russia
NETWORK RELIABILITY ASSESSMENT BY MONTE CARLO METHOD USING DECOMPOSITION BY ARTICULATION POINT
DOI: 10.24412/2073-0667-2026-2-32-42
EDN: OUEJBF
Analysis of network reliability is extremely important for their design and operation. For various types of networks, various models have been proposed that take into account network particular features, within which different indicators of network reliability are considered. As a rule, random graphs in various modifications are taken as a basis. Usually, the probability of connectivity of the corresponding random graph in the case of unreliable edges that fail independently and absolutely reliable nodes is used as an indicator of network reliability. The problems of exact calculating of various reliability indicators are NP-hard, so exact and fast estimation methods are needed for analysis of large scale networks.
The Monte Carlo method can be used as a universal method for estimating various reliability metrics. This method requires randomly generating a certain number of individual network realizations and averaging the desired metric across the resulting sample, which yields the desired reliability metric estimate. The corresponding results are presented in many well-known papers. In the paper, a new approach is presented.
The essence of Monte Carlo method for the problems under consideration consists of randomly generating a certain number of partial network realizations and averaging the metric of interest (in this case, connectivity) across the resulting sample, which serves as a reliability estimate.
To estimate the probability of graph connectivity, we use the connectivity indicator function, which is defined as a random variable on the space of elementary events. It takes two values: 0 and 1. The function is equal to 1 when event is successful, 0 otherwise. The success of event refers to the correct functioning of the corresponding network. In this paper, this is network connectivity.
This article proposes a new approach to reliability assessment using the Monte Carlo method: instead of generating a partial graph realization, we generate partial realizations of its components obtained by decomposing them by articulation point. Estimates of the number of trials for graph components to achieve the required solution error are presented. In practice, this can significantly speed up computations.
Key words: network reliability, random graph, biconncctcd graph, probabilistic connectivity, Monte Carlo method, network decomposition, articulation point, separator.
This work was carried out under state contract with ICMMG SB RAS FWNM-2025-0005.
References
1. Zhukovskij М. К.. Rajgorodskij A.M. Sluchajnve grafv: modeli i predel’nve harakteristiki // Uspekhi matematicheskih nauk. 2015. T. 70, N 1 (421). S. 35-88.
2. Mochalov V. A., Mochalova A.V. Primenenie ekspertnvh sistem diva rascheta verovatnosti svvaznosti mezhdu uzlami grafa // Gibridnve i sinergeticheskie intellektual’nye sistemv: materialv V Vserossiiskoi Pospelovskoi konferencii s mezhdunarodnym uchastiem. Pod red. A.V. Kolesnikova. 2019. S. 226-235.
3. Colbourn Ch. J. The combinatorics of network reliability. N. Y.: Oxford Univ. Press, 1987. P. 160.
4. Migov D. A. Formulv diva bvstrogo rascheta verovatnosti svvaznosti podmnozhestva vershin v grafah nebol’shoj razmernosti // Problemv informatiki. 2010. T. 6. S. 10-17.
5. Rodionov A. S. Mozhno li dobit’sva dal’nejshego uskoreniva rascheta harakteristik svvaznosti sluchajnogo grafa? // Problemv informatiki. 2022. N 4(57). S. 39-52.
6. Batenkov A. A., Batenkov K.A., Fokin A.B. Verovatnost’ svvaznosti telekommunikacionnoj seti na osnove privedeniva neskol’kih sobvtij nesvvaznosti к ob”edinenivu nezavisimvh sobvtij // Informacionno-upravlyavushchie sistemv. 2021. N 6. S. 53-63. DOI: 10.31799/1684-8853-2021-6-53-63.
7. Slvke R. V., Frank H. Network reliability analysis. Part I // Networks. 1972. V. 1. P. 279-290.
8. Cancela H., Urquhart M.E. Adapting RVR simulation techniques for residual connectedness network reliability models // IEEE Trans. Computers. 2002. V. 51, N 4. P. 439-443.
9. Botev Z.I., Kroese D.P. Efficient Monte Carlo simulation via the generalized splitting method // Statistics and Computing. 2012. V. 22, N 1. P. 1-16.
10. Migov D.A., Vins D.V. Parallel’nava realizaciva i imitacionnoe modelirovanie ocenki nadezhnosti seti metodom Monte-Karlo / / Vestnik Tomskogo gosudarstvennogo universiteta. Upravlenie, vvchislitel’nava tekhnika i informatika. 2019. N 47. S. 66-74. DOI: 10.17223/19988605/47/8.
11. Migov D. A. Dekompoziciva seti po sechenivam pri raschete ее nadezhnosti // Prikladnava diskretnava matematika. 2020. N 47. S. 62-86. DOI: 10.17223/20710410/47/6.
12. Burgos J.M. Factorization of network reliability with perfect nodes II: Connectivity matrix // Discrete Applied Mathematics. 2016. V. 198. P. 91-100
Bibliographic reference: Matsegora A. A., Migov D. A., Rodionov A. S. Network Reliability Assessment by Monte Carlo Method Using Decomposition by Articulation Point //"Problems of informatics", 2026, № 2, pp.32-42. DOI: 10.24412/2073-0667-2026-2- 32-42
Institute of Computational Mathematics and Mathematical Geophysics of SB RAS, 630090, Novosibirsk, Russia
ANALYSIS AND SYNTHESIS OF AN EXTENDED OPTIMAL DOUBLE-LOOP GRAPHS DATASET USING LLM
DOI: 10.24412/2073-0667-2026-2-43-58
EDX: QWGQYO
This paper is devoted to the actual problem of graph theory and network topology design: the search and analytical description of families of optimal two-dimensional ring circulant graphs. To solve this problem, we use a combination of analysis of the obtained extended dataset of optimal circulants, an original visualization method, and algorithmic search using large language models as a tool for generating program code and analyzing the dataset. The algorithm for synthesizing the dataset of optimal graphs is developed using large language models in Python and implemented in both sequential and parallel versions on the Kunpeng processor. The work expands the well-known database of optimal two-dimensional circulants and offers new families of graphs with analytical tasks. The paper describes the experience of working with large language models as a modern approach to automating scientific research.
Key words: two-dimensional double-loop graph, optimal graph, dataset, parametric description of circulant families, large language model.
Supported by state assignment of ICMMG SB RAS N FWNM-2025-0005.
References
1. Hwang F. K. A survey on multi-loop networks // Theoret. Computer Science. Vol. 299. 2003. P. 107-121. DOI: 10.1016/S0304-3975(01)00341-3.
2. Monakhova E. A. A Survey on Undirected Circulant Graphs // Discr. Math. Algorithms and Appl. Vol. 4. 2012. ID 1250002. 30 p. DOI: 10.1142/S1793830912500024.
3. Chen B.-X., Meng J.-X., Xiao W.-J. Some new optimal and suboptimal families of undirected double-loop networks // DMTCS. Vol. 8. 2006. P. 299-312.
4. Bermond J.-C., Tzvieli D. Minimal diameter double-loop networks: Dense optimal families // Networks. Vol. 21. 1991. P. 1-9.
5. Huang X., Ramos A. F., Deng Y. Optimal circulant graphs as low-latencv network topologies // J. Supercomput. Vol. 78. 2022. P. 13491-13510. DOI: 10.1007/sll227-022-04396-5.
6. Monakhova E.A., Monakhov O.G., Romanov A. Yu. Routing Algorithms in Optimal Degree Four Circulant Networks Based on Relative Addressing: Comparative Analysis for Networks-on- Chip // IEEE Transactions on Network Science and Engineering. Vol. 10. 2023. P. 413-425. DOI: 10.1109/TNSE.2022.3211985.
7. Liu H., Li X., Wang S. Construction of Dual Optimal Bidirectional Double-Loop Networks for Optimal Routing // Mathematics. Vol. 10. 2022. P. 4016.
8. Chen B.-X., Meng J.-X., Xiao W.-J. A constant time optimal routing algorithm for undirected double-loop networks // Proceedings of the First Int. Conf. Mobile Ad-hoc and Sensor Networks (MSN 2005), Wuhan, China, December 2005. LNCS 3794. P. 308-316.
10. Monakhova E.A., Monakhov O.G. Generaciva i analiz dataseta optimal’nvh dvuhkonturnvh cirkulvantnvh setej // Programmnava inzheneriva. Vol. 15(8). 2024. P. 402-410. DOI: 10.17223 20710110 17 7. (In Russian)
11. Monakhov O. G., Monakhova E. A. Masshtabiruemvj podhod к kodizajnu topologij i algoritmov marshrutizacii diva semejstv optimal’nyh cirkulvantnvh setej stepeni chetvre // Diskretnvj analiz i issledovanie operacij. Vol. 32, N 2. 2025. P. 88-106. (In Russian)
12. Tzvieli D. Minimal diameter double-loop networks. 1. Large infinite optimal families //
Networks. Vol. 21, N 4. 1991. P. 387-415.
13. Monakhova E. A., Monakhov O.G. Analiz bazv dannvh optimal’nyh dvuhkonturnvh kol’cevvh setej // Prikladnava diskretnava matematika. 2024. N 64. P. 56-71. (In Russian)
14. Chen B.-X., Meng J.-X., Xiao W.-J. A diameter formula for an undirected double-loop network // Ars Combinatoria. Vol. 90. 2009. P. 395-404.
15. Tzvieli D. Double-loop interconnection networks with minimal transmission delay: Ph.D. Dissertation. Louisiana State University, 1988.
16. Monakhova E. A. Optimal’nve cirkulvantnve grafv s prvamougol’nym konturom ukladki na ploskosti // Sibirskie elektronnve matematicheskie izvestiva. Vol. 22(2). 2025. P. 1538-1551. DOI: 10.33048/semi.2025.22.093. (In Russian)
17. Hwang F. K. A complementary survey on double-loop networks // Theoret. Computer Science. Vol. 263. 2001. P. 211-229
Bibliographic reference: Monakhova E.A., Monakhov 0. G. Analysis and Synthesis of an Extended Optimal Double¬Loop Graphs Dataset Using LLM //"Problems of informatics", 2026, № 2, pp..43-58 DOI: 10.24412/2073-0667-2026-2-43-58
N.E. Motornyj
Peter the Great St. Petersburg Polytechnic University, 195251, Saint-Petersburg, Russian Federation
EXPERIMENT ON MONITORING THE ORIGINALITY OF PROGRAM DEVELOPMENT THROUGH CODE EDITOR EVENTS ANALYSIS
DOI: 10.24412/2073-0667-2026-2-59-75
EDX: GKYTQO
The rapid advancement of generative Large Language Models (LLMs) has significantly simplified the completion of typical laboratory assignments for engineering students. Students can obtain ready-made solutions for any lab variant without intellectual effort by making a single query to an LLM chatbot.
Consequently, generating multiple variants and altering problem formulations have become almost completely inefficient against plagiarism. In-person debates with a professor still can significantly improve chances to detect assignments that were made using LLMs or direct plagiarism from other sources, however, in-person discussions arc not infinitely sealable when considering a student-to-teaher ratio because of a teacher’s time constraints. This paper practically addresses the challenge of ensuring independent completion of laboratory programs within the “Algorithms and Data Structures” course. The objective of this work is to conduct an experiment aimed at improving the process of monitoring originality of work completion by using sealable automated tools both for plagiarism monitoring and results control. The experiment is focused on implementing a system that analyzes the code writing process the student actions within a text editor, rather than evaluating solely the final source-code solution file. The authors developed and integrated a web-based editor based on Monaco Editor (Visual Studio Code) with an existing automatic grading system to record all text input events. During the experiment, 77 students worked on a total of 140 lab variants. Data were collected on code typing events from the text editor within the web interface, along with information on solution progress, the number of submission attempts, and copy attempts. Empirically selected threshold values for basic copy protections allowed blocking attempts to directly paste the entire code volume, thereby ensuring the quality and variability of the collected dataset. It is still clear that char-by-char copying of a premade work is possible within the provided constraints, however these basic checks worked as expected. The fastest assignment completion took about 20 minutes. In total, over 700 000 editing events were collected, with a total volume of modified characters exceeding 1.7 million.
The article presents both a general analysis of all student actions and an individual analysis of each of the 117 successfully completed assignments. A total of 2,699 submission attempts were made, of which 361 failed these plagiarism check. After passing the plagiarism check, the source code was submitted for compilation. The proportions of successful compilations for both laboratory assignments arc similar and approximate 50 %. Complied programs were then evaluated on a set of a static test- cases with time and memory limits. Successful pass of more than 80 % of test-cases was considered as assignment complitcion, which then formed a dataset of in-depth analyzed 117 submissions. The results provide insights into typing patterns, general statistical characteristics of assignment editor events timelines, and the relationship between task type and the problem-solving process. Analysis of
keystrokes counts, ratios between keystrokes and edited symbols count as well relation with submission success are presented. For the detailed analysis, such characteristic as cumulative sum of input symbols, input speed and its standard deviation are considered important. Other explored characteristics were: mean, median and standard deviation of keystroke event pauses, editing position both for column and line coordinates of the code-editor. Two different tasks were presented to students, the first required implementation of memory management module and contained slightly different variations for students, the second required search-tree implementation and the problem definitions were identical for all students. С programming language was the only allowed choice for submissions.
It is concluded that text editor event analysis is an effective tool for monitoring originality and lack of plagiarism in the assignments. The collected data volume is validated as correct and sufficient for the further development of advanced analysis methods, including those utilizing machine learning. The implementation of in-depth analysis will help to eliminate cases of char-by-char transcription from external sources (e.g., LLM chats) into the assessed code editor — such cases were detected during the presented experiment. The obtained results can be utilized for designing automated educational environments that are scalable in terms of the student-to-teacher ratio while preserving the quality of software engineer training.
Key words: laboratory assignments, Large Language Models (LLM), plagiarism detection, programming education, web-based editor, code-editor events analysis, work originality control.
References
1. Irtegov, D. V. Sistemv1 avtomatizirovannoj ocenki zadanij po programmirovanivu: razrabotka, ispokzovanie i perspektivv1 / D. V. Irtegov, Т. V. Nesterenko, T. G. Churina // Vestnik Novosibirskogo gosudarstvennogo universiteta. Seriva: Informacionnv'e texnologii. 2019. T. 17, N 2. S. 61-73. DOI 10.25205/1818-7900-2019-17-2-61-73. EDN CCAWIW.
2. Krasheninnikov S. V. Development and implementation of intellectual components of the digital educational environment / S. V. Krasheninnikov, D. A. Kurnosov, M. A. Nakhatovich, F. A. Novikov // EDULEARN23 Proceedings. Palma : IATED, 2023. P. 6535-6541.
3. Vinokurova D. V. Metod generacii unikakny'x variantov diva matematicheskix zadach / D. V. Vinokurova // Komp‘yuterny‘e instrumentv1 v obrazovanii. 2024. N 1. S. 71-84. DOI: 10.32603/2071¬2340-2024-1-100. EDN: ALXKFT.
4. Samigulina G. A. Obzor intellektuaknv'x sistem diva distancionnogo obrazovaniva, postroennv'x na osnove muktiagentnogo podxoda / G. A. Samigulina, A. T. Nvusupov // Problemv1 informatiki. 2017. N 2(35). S. 24-37. EDN: ZOLFNV.
5. Amaeva L. A. Sposob sbora dannv'x о vzaimodejstvii obuchavushhegosva s sistemoj edektronnogo obucheniva / L. A. Amaeva // Vestnik komp'yuternv'x i informacionnv'x texnologij. 2023. T. 20, N 5(227). S. 20-27. DOI: 10.14489/vkit.2023.05.pp.020-027. EDN: EDYBWP.
6. Ma D. Sistema s obratnoj svyaz'yu diva ocenki sostovaniva cheloveka po ego manere rabotv1 na klaviature / D. Ma, YU. A. Shichkina // Komp‘yuterny‘e instrumentv1 v obrazovanii. 2024. N 3. S. 66-78. DOI: 10.32603/2071-2340-2024-3-66-78. EDN: UNBUMP.
7. Gaines R. S. Authentication by keystroke timing: some preliminary results [Text] / R. S. Gaines, W. Lisowski, S. J. Press, N. Shapiro. Santa Monica : Rand Corporation, 1980. Report R-256-NSF.
8. Garcia J. M. A study of the feasibility of using keystroke dynamics for user identification [Text]: Ph.D. diss. / J. M. Garcia. Monterey: Naval Postgraduate School, 1986.
9. Wirth N. Program Development by Stepwise Refinement [Text] // Communications of the ACM. 1971. Vol. 14, N 4. P. 221-227.
10. van der Aalst W. M. P. Process Mining: Data Science in Action [Text]. 2nd ed. Heidelberg: Springer, 2016. 467 p. ISBN 978-3-662-49850-7.
11. Begicheva А. К. Discovering hierarchical process models: an approach based on events partitioning [Text] / A. K. Begicheva, I. A. Lomazova, R. A. Nesterov // Modeling and Analysis of Information Systems. 2024. Vol. 31. N 3. R 294-315. DOI 10.18255/1818-1015-2024-3-294-315. EDN LJPXWS.
12. Chen Z. Deep learning for keystroke dynamics: A comprehensive survey [Text] / Z. Chen, J. Zhang, Y. Zhang // ACM Computing Surveys. 2020. Vol. 53, N 4. Article 78. 35 p. DOI: 10.1145/3397029
Bibliographic reference: Motornyj N. E. Experiment on Monitoring the Originality of Program Development through Code Editor Events Analysis //"Problems of informatics", 2026, № 2, pp..___. DOI: 10.24412/2073-0667-2026-2-59-75
J. Rahmani, E.A. Karavaeva
Moscow Technical University of Communications and Informatics, 107045, Moscow, Russian Federation
A QUANTITATIVE MODULE CHANGE-COST MODEL FOR JUSTIFYING ARCHITECTURAL INTERVENTION IN MODULAR MONOLITHS
DOI: 10.24412/2073-0667-2026-2-76-96
EDX: DGVRLR
The paper considers the problem of quantifying architectural intervention and selecting candidate modules for service allocation within a modular monolith. A weighted multi-criteria cost function, Costj, is proposed, which combines four blocks: evolutionary load, post-release defect rate, median recovery time, and integration overhead. Empirical testing on 18 open enterprise-class projects with an explicit modular structure for 2019-2024 confirmed: module ranking is resistant to weight profile changes (Kendall t 0.75-0.97); multi-criteria function reduces the number of false alarms of single-factor metrics; projects with formalized architectural 1' units have a median defect load of 0.283 versus 0.443 for informal ones (p = 0.0416, Cohen's d = 0.89). The method relies solely on public repository data and can be reproduced on any open-source project with an accessible change history. It can be transferred to proprietary systems if comparable repository and incident data is available.
Key words: modular monolith, microservice architecture, change cost, multi-criteria evaluation, SAW, TOPSIS, repository metrics, co-change.
References
1. Lohman D., Wacke M. Scaling up the Prime Video audio/video monitoring service and reducing costs by 90% // ACM Queue. 2023. Vol. 21, N 2. (Electron. Res.]
https://queue.acm.org/detail. cfm?id=3595833 (accessed: 01.03.2026).
2. Rahmani J., Dubrovskaya A. A., Lobanova A. A. Modul’nyi podkhod v razrabotke informatsionnvkh sistem: preimushchestva i realizatsiva (Modular approach in information system development: advantages and implementation] // Trudy Instituta sistemnogo analiza Rossiiskoi akademii nauk (Proceedings of the Institute for Systems Analysis of the Russian Academy of Sciences]. 2025. Vol. 75, N 3. P. 61, 67 (In Russian).
3. Radostev D.K., Nikitina E.Yu. Strategiva migratsii programmnogo koda iz monolitnoi arkhitekturv v mikroservisv (Migration strategy from monolithic to microservice architecture] // Vestnik Permskogo universiteta. Seriva: Matematika. Mekhanika. Informatika (Perm University Herald. Mathematics. Mechanics. Computer Science]. 2021. N 2(53). P. 65, 68 (In Russian).
4. Ziborev A.V. Antipatterv postroeniva mikroservisnvkh prilozhenii v vvsokonagruzhennvkh proektakh (Anti-patterns for microservice application design in high-load projects] // Universum: tekhnicheskie nauki (Universum: Technical Sciences]. 2023. N 11(116) (In Russian).
5. Chernikov B.V., Poklonov B.E. Otsenka kachestva programmnogo obespecheniva: praktikum (Software quality assessment: a practical course] / ed. by В. V. Chernikov. Moscow: Forum: INFRA-M,
2022. 400 p. ISBN 978-5-8199-0516-6 (In Russian).
6. Blinowski G., Ojdowska A., Przvbvlek A. Monolithic vs. Microservice Architecture: A Performance and Scalability Evaluation // IEEE Access. 2022. Vol. 10. P. 20357-20374. DOI: 10.1109/ACCESS.2022.3152803.
7. Parnas D. L. On the criteria to be used in decomposing systems into modules // Communications of the ACM. 1972. Vol. 15, N 12. P. 1053-1058. DOI: 10.1145/361598.361623.
8. Evans E. Domain-Driven Design: Tackling Complexity in the Heart of Software. Addison-Weslev, 2003. 529 p.
9. Newman S. Building Microservices. 2nd ed. O’Reilly Media, 2021. 620 p.
10. Nagappan N., Ball T. Use of relative code churn measures to predict system defect density // ICSE 2005. P. 284-292. DOI: 10.1145/1062455.1062514.
11. D’Ambros М., Lanza М., Robbes R. An extensive comparison of bug prediction approaches // MSR 2010. P. 31-41.
12. Bavota G. et al. How the Apache community upgrades dependencies: an evolutionary study // Empirical Software Engineering. 2015. Vol. 20. P. 1275-1317. DOI: 10.1007/sl0664-014-9308-x.
13. Zimmermann T. et al. Predicting defects for Eclipse // PROMISE 2007. IEEE, 2007.
14. Forsgren N., Humble J., Kim G. Accelerate: The Science of Lean Software and DevOps. IT Revolution Press, 2018. 288 p.
15. Taibi D., Lenarduzzi V., Pahl C. Processes, motivations and issues for migrating to microservices architectures // IEEE Cloud Computing. 2017. Vol. 4, N 5. P. 22-32. DOI: 10.1109/MCC.2017.4250931.
16. Hwang C.-L., Yoon K. Multiple Attribute Decision Making. Springer, 1981. 259 p.
17. Auer F., Lenarduzzi V., Felderer М., Taibi D. From monolithic systems to microservices: An assessment framework // Information and Software Technology. 2021. Vol. 137. Article 106600. DOI: 10.1016/j.infsof.2021.106600.
18. Chernikov B.V. Upravlenie kachestvom programmnogo obespecheniva: uchebnik [Software quality management: textbook]. Moscow: Forum: INFRA-M, 2022. 240 p. ISBN 978-5-8199-0902-7 (In Russian).
19. Kotlvarov V. P. Quality of software testing and traceabilitv // Programming and Computer Software. 2020. Vol. 46, N 5. P. 297-306. DOI: 10.1134/S0361768820050059.
20. Cheremnvkh O.N., Cheremnvkh S.V. Strategicheskii korporativnvi reinzhiniring: protsessno- stoimostnoi podkhod к upravlenivu biznesom [Strategic corporate reengineering: process-cost approach to business management]. Moscow: Finansv i statistika, 2005. 736 p. ISBN 978-5-279-02819-2 (In Russian).
21. Martin R. C. Agile Software Development: Principles, Patterns, and Practices. Prentice Hall, 2002. 529 р.
Bibliographic reference: Rahmani J., Karavaeva E. A. A Quantitative Module Change-Cost Model for Justifying Architectural Intervention in Modular Monoliths //"Problems of informatics", 2026, № 2, pp..76-96. DOI: 10.24412/2073-0667-2026-2-76-96
