Received 22.12.2024, Revised 18.03.2025, Accepted 24.04.2025

Analysis of integrated real-time decision support systems based on neural networks and low-structured data

Mykola Demchyna

The study aimed to analyse and substantiate effective methods for analysing inefficiently structured data using neural networks to provide operational decision support in complex environments. The focus was on the use of artificial neural networks to analyse inefficiently structured data, such as sensor streams, to ensure efficiency, accuracy and adaptability in a dynamic environment. The research is aimed at creating innovative models and technologies that will improve the efficiency of management in complex situations, such as emergency response, process automation in critical industries and decision-making based on predictive analytics. The study investigated conceptual approaches to the development of integrated real-time decision support systems based on the analysis of poorly structured data using neural networks. The study proposed methods of adaptive learning that allow neural networks to process data efficiently in the face of constant changes. The research methodology included modelling a real-time architecture using a microservice approach and streaming data processing platforms such as Apache Kafka and Apache Flink. The study highlighted the role of neural networks in processing streaming data, in particular, convolutional networks for processing visual information, recurrent networks for sequence analysis, and transformers for multichannel analysis. Architectural solutions were developed that allow the processing of large amounts of data with minimal delays, ensuring the accuracy and adaptability of systems. The study presented approaches to the implementation of adaptive training of neural networks that minimise the risks of losing model relevance in a dynamic environment. The use of modern technologies, such as artificial neural networks, adaptive learning and integration with the Internet of Things, was used to create effective systems for rapid response to emergencies. The proposed methods help increase the efficiency of management in difficult conditions and create new prospects for innovation in various industries

artificial intelligence; knowledge base; dynamic environment; network models; knowledge extraction
20-29
Demchyna, M. (2025). Analysis of integrated real-time decision support systems based on neural networks and low-structured data. Information Technologies and Computer Engineering, 22(1), 20-29. https://doi.org/10.63341/vitce/1.2025.20

References

[1] Abirami, S., & Chitra, P. (2020). Energy-efficient edge based real-time healthcare support system. Advances in Computers, 117(1), 339-368. doi: 10.1016/bs.adcom.2019.09.007.

[2] Apache Flink. (n.d.). Retrieved from https://flink.apache.org/.

[3] Apache Kafka. (n.d.). Retrieved from https://kafka.apache.org/.

[4] Ashraf, S., Afify, Y.M., & Ismail, R. (2022). Big data for real-time processing on streaming data: state-of-the-art and future challenges. In 2022 International conference on electrical, computer, communications and mechatronics engineering (pp. 1-8). Maldives: IEEE. doi: 10.1109/iceccme55909.2022.9987770.

[5] Deepthi, B.G., Rani, K.S., Krishna, P.V., & Saritha, V. (2023). An efficient architecture for processing real-time traffic data streams using Apache Flink. Multimedia Tools and Applications, 83(13), 37369-37385. doi: 10.1007/s11042-023-17151-6.

[6] Dhruv, P., & Naskar, S. (2020). Image classification using convolutional neural network (CNN) and recurrent neural network (RNN): A review. In D. Swain, P. Pattnaik & P. Gupta (Eds.), Machine learning and information processing (pp. 367-381). Singapore: Springer. doi: 10.1007/978-981-15-1884-3_34.

[7] Fan, F., Xiong, J., Li, M., & Wang, G. (2021). On interpretability of artificial neural networks: A survey. IEEE Transactions on Radiation and Plasma Medical Sciences, 5(6), 741-760. doi: 10.1109/trpms.2021.3066428.

[8] Gurcan, F., & Berigel, M. (2018). Real-time processing of big data streams: Lifecycle, tools, tasks, and challenges. In 2018 2nd international symposium on multidisciplinary studies and innovative technologies (pp. 1-6). Ankara: IEEE. doi: 10.1109/ismsit.2018.8567061.

[9] Haidur, H.I., Gakhov, S.O., & Bryhynets, A.A. (2023). Detection of network anomalies with neural networks algorithms. Telecommunication and Information Technologies, 1(78), 61-73. doi: 10.31673/2412-4338.2023.016173.

[10] Hammou, B.A., Lahcen, A.A., & Mouline, S. (2019). Towards a real-time processing framework based on improved distributed recurrent neural network variants with fastText for social big data analytics. Information Processing & Management, 57(1), article number 102122. doi: 10.1016/j.ipm.2019.102122.

[11] Han, Y., Huang, G., Song, S., Yang, L., Wang, H., & Wang, Y. (2021). Dynamic neural networks: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(11), 7436-7456. doi: 10.1109/tpami.2021.3117837.

[12] Hancock, J.T., & Khoshgoftaar, T.M. (2020). Survey on categorical data for neural networks. Journal of Big Data, 7, article number 28. doi: 10.1186/s40537-020-00305-w.

[13] Hariri, R.H., Fredericks, E.M., & Bowers, K.M. (2019). Uncertainty in big data analytics: Survey, opportunities, and challenges. Journal of Big Data, 6(1), article number 44. doi: 10.1186/s40537-019-0206-3.

[14] Huang, G., Zhu, Q., & Siew, C. (2006). Real-time learning capability of neural networks. IEEE Transactions on Neural Networks, 17(4), 863-878. doi: 10.1109/tnn.2006.875974.

[15] Huang, S., Lin, C., Zhou, K., Yao, Y., Lu, H., & Zhu, F. (2020). Identifying physical-layer attacks for IoT security: An automatic modulation classification approach using multi-module fusion neural network. Physical Communication, 43, article number 101180. doi: 10.1016/j.phycom.2020.101180.

[16] Kabudi, T., Pappas, I., & Olsen, D.H. (2021). AI-enabled adaptive learning systems: A systematic mapping of the literature. Computers and Education Artificial Intelligence, 2, article number 100017. doi: 10.1016/j.caeai.2021.100017.

[17] Kavitha, D., & Ravikumar, S. (2020). IOT and context‐aware learning‐based optimal neural network model for realtime health monitoring. Transactions on Emerging Telecommunications Technologies, 32(1), article number e4132. doi: 10.1002/ett.4132.

[18] Liu, Y., He, Q., Zheng, D., Xia, X., Chen, F., & Zhang, B. (2020). Data caching optimization in the edge computing environment. IEEE Transactions on Services Computing, 15(4), 2074-2085. doi: 10.1109/tsc.2020.3032724.

[19] Mehmood, E., & Anees, T. (2020). Challenges and solutions for processing real-time big data stream: A systematic literature review. IEEE Access, 8, 119123-119143. doi: 10.1109/access.2020.3005268.

[20] Mohammadi, M., Al-Fuqaha, A., Sorour, S., & Guizani, M. 2018. Deep learning for IoT big data and streaming analytics: A survey. IEEE Communications Surveys & Tutorials, 20(4), 2923-2960. doi: 10.1109/COMST.2018.2844341.

[21] Mohan, D., & Thyagarajan, K. (2023). A side-by-side comparison of Apache Spark and Apache Flink for common streaming use cases. Retrieved from https://aws.amazon.com/ru/blogs/big-data/a-side-by-side-comparison-of-apache-sparkand-apache-flink-for-common-streaming-use-cases/?utm_source.