European Journal of Science, Innovation and Technology

Adapting Code Review Processes to the Conditions of AI-Assisted Software Development

2026-05-06T10:07:17+03:00

The article presents an analysis of the transformation of code review processes in the context of software development with the use of artificial intelligence. The study is conducted in the format of a systematic review and analytical synthesis of academic publications focusing on code generation, development productivity, and the organization of change validation processes. The primary focus is on the relationship between the speed of code creation, the characteristics of introduced changes, and the ability of the review process to ensure their reliable interpretation and safe integration. The key parameters determining the effectiveness of code review are examined, including temporal characteristics, the structure of changes, and cognitive load. It is established that the impact of code generation tools is indirect and manifests through the formation of asymmetry between the intensity of generation and the capacity for review. It is shown that even with improvements in the quality of automatically generated code, the risk of uncritical acceptance and the integration of hidden logical inconsistencies remains. An original adaptive code review model is proposed, reflecting a transition from single-stage review to a multi-level system that includes preliminary filtering, structured analysis, risk assessment, and a feedback loop. The results obtained allow code review to be considered as a controllable risk regulation mechanism determining the stability and predictability of the development process. The practical significance of the study lies in substantiating the need to structure code review processes, limit the scope of changes, and increase transparency in the origin of code fragments. The article will be useful for software developers, engineering managers, and specialists in software process management.

Molecular Detection and Statistical Association of Carbapenem Resistance Determinants (blaOXA and blaNDM) in Clinical Isolates of Escherichia coli and Klebsiella pneumoniae from Southeastern Nigeria

2026-05-11T21:16:01+03:00

Background: Carbapenem-resistant Enterobacteriaceae (CRE) represent a critical public health threat, with limited epidemiological data from southeastern Nigeria. This study investigated the prevalence of carbapenem resistance and the distribution of blaOXA and blaNDM genes among clinical isolates of Escherichia coli and Klebsiella pneumoniae from a tertiary hospital in Enugu, Enugu State.

Methods: A total of 210 non-duplicate urine samples were collected from patients attending Enugu State University Teaching Hospital (ESUTH). Bacterial isolates were identified using standard microbiological techniques and API 20E system. Antimicrobial susceptibility testing was performed by disk diffusion method, with carbapenem resistance phenotypically confirmed using the Modified Hodge Test. Molecular detection of blaOXA and blaNDM genes was conducted using conventional PCR. Statistical associations between resistance phenotypes and demographic variables were analyzed using SPSS version 25.

Results: Gram-negative bacteria were isolated from 151 (71.9%) samples, comprising 89 (42.4%) E. coli and 62 (29.5%) K. pneumoniae. Phenotypic carbapenem resistance was detected in 24 (15.9%) isolates: 10 (11.2%) E. coli and 14 (22.6%) K. pneumoniae. All carbapenem-resistant isolates exhibited multidrug resistance phenotypes. Molecular analysis revealed that 21 (87.5%) of phenotypically resistant isolates harbored carbapenemase genes: blaNDM was detected in 12 (50.0%) isolates, blaOXA in 5 (20.8%), and co-occurrence of both genes in 4 (16.7%). K. pneumoniae demonstrated significantly higher gene carriage (92.9%) compared to E. coli (80.0%). Statistical analysis showed significant association between carbapenem resistance and patient age >40 years (p=0.023), but not with gender (p=0.412).

Conclusion: This study documents a high prevalence of blaNDM and blaOXA genes among CRE isolates in southeastern Nigeria, with K. pneumoniae serving as a major resistance reservoir. The findings underscore the urgent need for enhanced surveillance, antimicrobial stewardship, and infection control measures in the region.

Factors Influencing Green Logistics Adoption Intentions among Small and Medium Enterprises (SMEs) in Vietnam: An Empirical Study Based on the TOE Framework

2026-05-12T10:36:26+03:00

In an era of global environmental volatility, transitioning to sustainable practices has become a strategic imperative for Small and Medium Enterprises (SMEs) within the global value chain. This study investigates the factors influencing green logistics adoption intentions among SMEs in Vietnam by integrating the Technology-Organization-Environment (TOE) framework. Data were collected through a quantitative survey of 367 managers and owners of Vietnamese SMEs and analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM) via SmartPLS 4. The empirical results indicate that all seven proposed hypotheses are supported. Notably, top management support and regulatory pressure emerged as the most significant predictors of adoption intention. Furthermore, while relative advantage and compatibility positively influence adoption, technological complexity remains a critical barrier. Market pressure and the quality of green human resources also play vital roles in driving the green transition. Theoretically, this research extends the applicability of the TOE framework to the SME context in an emerging market. Practically, the findings offer evidence-based insights for policymakers to design targeted support mechanisms and for SME managers to strategically align their operations with international environmental standards.

Diffusion and Adsorption Characteristics of Crude Oil-Contaminated Silty Clay Soil Stabilised with Freshwater Limpet Shell Ash and Cement Combination

2026-05-13T20:04:19+03:00

The diffusion of Total Petroleum Hydrocarbon (TPH) pollutants to groundwater at crude oil-contaminated sites in Nigeria's Niger Delta is a problem that severely affects drinking water. But the inclusion of cementitious materials will reduce the diffusion rate to groundwater. This study analyses the Effective Diffusion Coefficient (EDC)/Tortuosity Factor (TF) and Retardation Factor (Rd) of Crude Oil-Contaminated Soils (COCS) stabilised with Freshwater limpet shell ash and cement combination (FWLSA-C). Uncontaminated soils were sampled at a depth of 1.0 m and mixed with 0, 2, 5, 10, 15 and 20% of crude oil by weight to produce artificially contaminated soils. Freshwater limpet shells were sourced from the Esaman River, Delta State, calcined to ash at 887°C for four hours and ground to powder. The ash oxide content was determined with X-ray fluorescence. The soil samples were mixed with different amounts of cement and ash. The combinations were: 0% cement and 0% ash, 5% cement with 2% ash, 5% cement with 3% ash, 5% cement with 5% ash, 5% cement with 10% ash, 5% cement with 15% ash, and 5% cement with 20% ash. The EDC/TF and Rd were determined using standard methods. The EDC/TF and Rd for 0% COCS without FWLSA-C were 0.00 m²/s/0.00 and 0.00; 20% COCS without FWLSA-C were 2.06E-10 m²/s/7.09E-02 and 11.64; and the average of 9.31E-11 m²/s/6.92E-02 and 28.49 by adding the FWLSA-C. The FWLSA-C reduced the EDC/TF from 2.06E-10 to 3.50E-11 m²/s/7.09E-02 to 6.66E-02 and increased Rd from 11.64 to 40.16, respectively. Therefore, this study revealed that freshwater limpet shell ash and cement combination (FWLSA-C) will slow the diffusion rate of Total Petroleum Hydrocarbon (TPH) contaminants to groundwater in crude oil-contaminated sites within the Niger Delta in Nigeria.

Data-Driven Assessment of Low-Speed Wind Energy Harvesting Potential Using MATLAB/Simulink Modeling

2026-05-13T20:18:09+03:00

This study presents a data-driven evaluation of a wind energy harvesting system designed for low power applications, particularly wireless sensor networks (WSNs). High resolution wind speed data were collected over a one-year period from January-December 2025 at the Federal University of Petroleum Resources, Effurun (FUPRE), Nigeria, using a calibrated Automated Weather Observing Station (AWOS). A comprehensive system model was developed in MATLAB/Simulink, incorporating aerodynamic wind turbine dynamics, electromagnetic energy conversion, rectification, DC-DC conversion, maximum power point tracking (MPPT) using the perturb and observe (P&O) algorithm, and a power management system (PMS) with battery storage. Simulation results reveal strong seasonal variations in harvested power, with peak performance observed in March, where average DC-DC output reaches approximately 38.022 mW and daily harvested energy peaks at about 0.766 Wh/day. Lower performance is recorded during periods of reduced wind activity, particularly between October and November. Despite these fluctuations, the PMS effectively regulates power flow, ensuring stable energy delivery to the load. The system demonstrates a relatively constant efficiency of approximately 16.47% across all months, indicating stable internal conversion performance. Overall, the results confirm the feasibility of utilizing low-speed wind energy for continuous power supply in autonomous systems. The study highlights the effectiveness of integrating real environmental data with intelligent power management strategies to enhance reliability and energy utilization in wind energy harvesting applications.

AI-Driven Spatial Biomarker Discovery in the Tumor Microenvironment

2026-05-18T21:26:08+03:00

How a cancer develops and how well it responds to treatment isn't just about the cancer cells themselves anymore; the area around the tumour (the tumour microenvironment) is now understood to be hugely important, and this includes how the cells are organised within the tissue. As we’ve learned more about the location of cells in a tissue from recent developments in spatial biology, we've seen that where immune cells and cancer cells are in relation to each other, and how they interact, greatly affects how the disease will go (Erasha et al., 2025; Williams et al., 2024). Spatial biomarkers - which are defined by where cells are, how many of each type are in an area, and how they’re positioned next to each other - give a much more detailed and accurate prediction of what will happen compared to the usual ‘bulk’ biomarkers (Faktor et al., 2024).

Using artificial intelligence (AI) with this spatial analysis has really sped up the finding of biomarkers, because it’s able to pick out complicated patterns from a lot of information from images and what genes are being used. Machine learning and deep learning methods help us to find spatial patterns linked to immune cell movement into the tumour, how different the tumour cells are from each other, and resistance to treatments (Nguyen et al., 2021; Xu et al., 2024). Spatial transcriptomics and looking at digital pathology images are the foundations for this, providing the data for AI to work out what cells are near which others and what those relationships mean for biology (Du et al., 2023; Jin et al., 2024).

What’s particularly exciting is that using AI to find these spatial biomarkers is showing great promise for predicting whether a treatment will work, especially with immunotherapies. We can now categorize tumours as either “hot” or “cold” for immune activity, and we can find patterns of keeping immune cells out of the tumour. These both improve how we decide which patients to treat in which ways (Melssen et al., tually, and therapeutic decisions (Melssen et al., 2023; Zuo et al., 2020). Moreover, newer computer programs that use graph-based methods and combine multiple kinds of data are making the predictions even more accurate and useful in the clinic (Kong et al., 2021; Mallya et al., 2025).

Even with all this progress, difficulties still exist: the data can be quite variable, it can be tricky to understand how the AI models are working, and we definitely need solid tests in the clinic to confirm the findings. However, the coming together of AI and spatial biology is completely changing cancer research, providing a level of understanding of how the tumour and the immune system interact that we haven't had before, and opening the door for treatment plans designed for each individual's specific tumour's spatial characteristics.

Voice and Visual Framing in Television News: Influence of Narration on Viewer Perception in Kazakhstan

2026-05-19T20:26:04+03:00

The paper examines the alliance of voice narration in concert with the visual framing of televised news and how that relationship contributes to the perceptions of audiences who view televised news in the post-Soviet media space of Kazakhstan, which has experienced both the effects of state influence and media pluralism. Reflecting the elements of framing theory, narrative paradigm, visual semiotics as well as the aspects of receiving theory, the study establishes the influence of frames and codes of tones, pace, and the language used in voice-overs together with symbolic images and visual patterns over the trust of audience, which affects the emotional attachment and political sense. Using the convergent approach to a mixed-method study, data were analyzed based on 30 news items of large Kazakhstani broadcasters and focus groups were completed with a focus group made up of 24 interviewees of all age groups, languages, and geographical backgrounds. The results indicate the predominance of authoritative male voice and nationalistic imagery in entrenching state legitimacy and social harmony discourse. Although the target audiences, including older, rural, and Kazakh-speaking viewers, generally accepted these frames, younger, urban, and Russian-speaking readers were less compliant with them and frequently negotiated or challenged the messages that were supposed to be perceived. Such outcomes indicate the context-sensitive, two-fold character of media framing and acceptance in transitional democracies, as it may be associated with media ethics, literacy, and stance on the state-controlled news context.

Species Identification of Bushmeat Marketed by Using DNA Barcoding in the Region of Kisangani, D. R. Congo

2026-05-22T20:31:33+03:00

The Congo Basin Rainforest is characterized by a rich diversity in mammals. Scientific research demonstrated that this biodiversity is currently under high pressure. The survival of the fauna is threatened due to overexploitation as the demand for bushmeat as the primary source of animal protein increases due to the growing human population. This phenomenon is observed in the region of Kisangani, the biggest city in the northeast of the Democratic Republic of Congo (DRC).

Method: This study reports on the species diversity involved in the bushmeat trade in the region of Kisangani by means of DNA barcoding of mitochondrial DNA fragments cytochrome c oxidase subunit I (COI) and cytochrome b. An identification success rate of 65% was achieved (231 of 353 samples).

Research showed that in fact all mammal species are hunted, including endangered species as the chimpanzee (Pan troglodytes) and the okapi (Okapia johnstoni). Duikers (Cephalophinae) are the most important group of bushmeat species in terms of numbers (34%). Mammal species, except for rodents, become depleted in the close surroundings of Kisangani due to overexploitation. Mainly big-sized and medium-sized animals, which are profitable enough to compensate for the transport costs, are available on the market in Kisangani. Animals larger than 40 à 45 centimeters (shoulder height) are killed with shotguns, smaller animals are trapped. The bushmeat trade is explored more into detail by interviewing all parties involved in the bushmeat trade from hunters over sellers to the consumers, with regard to conservation of the fauna.

This dissertation proved overexploitation of wild animals in the region of Kisangani, DRC. Bushmeat consumption is reduced most effectively when it is becoming more expensive than alternative protein resources, rather than increasing the awareness on overexploitation or diseases emerging from wild animal consumption.

Characterization and Monitoring of Wildlife Trafficking at the Bayangana Checkpoint (Kisangani, DRC)

2026-05-22T20:34:33+03:00

In the Democratic Republic of the Congo, hunting and the sale of game are essential for feeding families and providing a livelihood for many people. But this activity also threatens the country’s rich wildlife. The Bayangana checkpoint, at the entrance to Kisangani, is a key location for observing what is really happening: every day, we see the game harvested from the surrounding forests passing through here.

To better understand this, a study was conducted specifically on this issue. We recorded which animals were being transported, distinguishing between authorized species and protected species. We weighed the shipments whether fresh, smoked, or dried and traced the journey of this meat from the forest to the town.

Observations show that hunting particularly of small mammals such as antelopes, monkeys, and rodents is widespread. This activity peaks during certain seasons, often coinciding with harvest times or when hunting is easier. Most concerning is that strictly protected species continue to be poached and sold, highlighting just how difficult it is to enforce the laws.

Given this situation, it has become urgent to better control access to the city and to help communities find alternative sources of food and income. The stakes are high: we must both preserve the forest and its wildlife and enable people to live with dignity. Striking this balance is the major challenge for the future.

Diversity of Primates in the Mukingiti and Kingombe Gorilla Reserve in Lubutu Territory, Bitule Sector (Maniema Province, D.R. Congo)

2026-05-22T20:37:40+03:00

Our study focused on the Mukingiti Gorilla Reserve (REGOMUKI), a buffer zone of Maiko National Park in the Democratic Republic of the Congo. This reserve is managed directly by local communities through the Bitule and Omate Local Community Forest Concessions (CFCL). However, increasing human pressure is observed, including poaching, agricultural clearing, and mining activities. Despite this, the area has maintained remarkable ecological integrity.

Our field inventories recorded no fewer than nine primate species. These include the Chimpanzee and Grauer’s Gorilla, an endemic species of the region that is critically threatened. Chimpanzees were the most frequently observed. In contrast, colobus monkeys, which are more sensitive to hunting pressure, are becoming increasingly rare.

This forest, with its high species diversity and apparently stable populations, proves to be an essential refuge. It undoubtedly constitutes a true sanctuary, a biodiversity center of exceptional value that must be preserved.

Distribution and Structure of Mammal Communities within Unclassified Forest Areas in the Province of Tshopo (DR Congo)

2026-05-22T20:41:17+03:00

The study on the distribution and structure of mammal communities within unclassified forest areas in the province of Tshopo (DR Congo) focuses on how medium to large sized mammal communities are organized. We seek to understand how wildlife is distributed in these areas, which are not officially protected and are subject to various human pressures. We have implemented an inventory protocol based on a reconnaissance method. To ensure the reliability of our data, the sampling effort was validated by calculating the completeness of the inventories, demonstrating capture rates of over 90% at all three sites, which guarantees the scientific representativeness of the data collected.

The structure of communities was analyzed using multivariate methods, including nonmetric multidimensional scaling (NMDS) to compare similarity between sites, and the calculation of specialization residuals helped identify the ecological preferences of each species. The spatial architecture of mammalian biodiversity reveals a biological organization of striking regularity. Ten species form the common base, representing 43% of the overall assemblage. The three sites share a core of generalist species and a major ecological gradient, with each site presenting signature and specialist species. Beta diversity ensures regional gamma richness, and species dominance differs according to local ecological constraints.

In conclusion, this study shows that despite the proximity of National Road No. 4, these forest areas still harbor structured and diverse animal communities. This highlights how crucial these areas, although unclassified, are for the preservation of biodiversity in the Democratic Republic of Congo.

Faunal Composition of Medium- and Large-Sized Mammals in Unclassified Forests along National Highway 4 in Tshopo Province (Democratic Republic of the Congo)

2026-05-22T20:44:54+03:00

The aim of this study was to assess the composition, diversity, and abundance of mammal communities at the three sites (Baboro/Libombo, Bafwakondima, and Dikwa).

Field data were collected using the “Recce” reconnaissance walk method, and the analyses focused on species richness, diversity indices, Pielou’s evenness index, the kilometer abundance index (KAI), the Morisita index, and the analysis of indicator species (IndVal).

A total of 2,028 indices belonging to several taxonomic groups were recorded, with artiodactyls being the most dominant across all sites. Bafwakondima exhibited the greatest species richness (17 species), the highest diversity (H’=2.32), and the greatest evenness (0.82), indicating a more balanced and ecologically stable community. In contrast, Baboro/Libombo and Dikwa exhibited lower diversity and a more pronounced dominance of certain species.

The Morisita index revealed a high similarity between Baboro/Libombo and Dikwa (0.96), while Bafwakondima stood out (0.71). The analysis of indicator species identified several significant species at Bafwakondima, highlighting its ecological importance.

Human activities, including hunting, logging, and mining, have been observed at all sites and are likely to influence community structure and species distribution.