No. 21/2025

Online archive of Computer Science and Mathematical Modelling

No. 21/2025

1. Kalutara Koralalage Shalidi Nadishani Britto, Anuradha P. Hewaarachchi - Modelling maximum and minimum temperature series in Trincomalee District, Sri Lanka

   Pages: 5 - 9
   Abstract: Sri Lanka is a tropical island in the Indian Ocean, which has a diverse climate shaped by monsoonal variations and seasonal fluctuations that have a significant impact on temperature trends, underscoring the need for accurate forecasting models. This study focuses on Trincomalee District, one of the hottest regions in Sri Lanka, where the highest recorded temperature from 1949 to April 2024 was 39.5°C. The research aims to model and forecast the monthly mean maximum and minimum temperature series using Periodic Autoregressive (PAR) and Seasonal Autoregressive Integrated Moving Average (SARIMA) models. Monthly data from January 2008 to October 2024 was analyzed to assess model performance in capturing seasonal and periodic trends. The SARIMA (1, 0, 0) (2, 1, 1) model best fits the maximum temperature series, with a low RMSE (0.7830) and MAPE (1.7628%). Meanwhile, the minimum temperature series shows significant periodicity, with the PAR(1) model offering better precision (RMSE 0.4391, MAPE 1.4071%). These findings support to improvement of climate forecasting and agricultural planning in Sri Lanka’s dry zone.
   Keywords periodic correlations, seasonal ARIMA, temperature
   Full article: 5_9_britto_hewaarachchi_modelling_max_min_temp_csmm_21_2025.pdf
2. Mateusz Maj - Performance comparison of Julia and Python programming languages on GPU and CPU

   Pages: 11 - 23
   Abstract: The paper presents research on the time performance of each language when performing operations on the CPU (central processing unit) and GPU (graphics processing unit).This article should be considered as a coherent continuation of the first part. The paper has been divided due to existing editorial criteria. The work includes descriptions of the technologies used and the algorithms that served the research purposes. Algorithms prepared with care to maintain equal use of functions by both programming languages based on block diagrams and pseudocode are presented. The paper first presents the research carried out on the basis of sorting algorithms implemented by the Author and executed on the processor. The next part of the paper presents research verifying the performance of Julia and Python programming languages on graphics cards, with a special focus on the implementation of matrix operations. The article is concluded with a discussion of the time performance of each programming language.
   Keywords Python, Julia, comparison, performance
   Full article: 11_23_mmaj_performance_comparison_csmm_21_2025.pdf
3. Paweł Kabata - Analysis of Vulnerabilities and Consequences of Missing Hardening in WordPress-Based CMS Environments

   Pages: 25 - 32
   Abstract: WordPress is currently the most widely used content management system and constitutes a key component of the modern web application ecosystem. Its evolution from a blogging platform into a universal CMS has significantly expanded its functionality while simultaneously increasing architectural complexity and the attack surface. The open-source nature of the project, the extensive ecosystem of plugins and themes, and a simplified permission model make the security of WordPress installations largely dependent on the quality of external components and the configuration of the hosting environment. This paper provides a review of the most common vulnerabilities observed in WordPress-based environments and discusses hardening techniques including update management, permission restrictions, authentication strengthening, and security monitoring. The analysis indicates that the absence of hardening measures leads to an increased risk of unauthorized access, malware infections, and operational and reputational losses, confirming the necessity of treating WordPress hardening as a continuous process.
   Keywords WordPress security, CMS, hardening
   Full article: 25_32_pkabata_artanalysis_csmm_21_2025.pdf
4. Norbert Waszkowiak - Runtime repository generation in Spring Boot: Advancing REST API automation with ByteBuddy

   Pages: 33 - 42
   Abstract: Spring Boot simplifies REST API development but still requires repetitive code for controllers, services, and repositories. This article presents a proof-of-concept for the @RestEntity annotation, which uses ByteBuddy to dynamically generate repositories at runtime. By utilizing Java reflection and bytecode manipulation, our approach automatically creates repository interfaces that Spring Boot seamlessly integrates into its context, significantly reducing boilerplate code while maintaining framework compatibility.
   Keywords code generation, spring boot, reflection
   Full article: 33_42_nwaszkowiak_csmm_21_2025.pdf