Journal of Knowledge in Data Science and Information Management

Evaluating the Effectiveness of AI-powered Code Review Tools in Improving Software Quality and Developer Productivity

Mission Franklin — 2026-01-20

The accelerating complexity of modern software systems, combined with the industry-wide demand for rapid release cycles, has increased the importance of efficient and high-quality code review processes. Traditional manual reviews remain vital for ensuring correctness and maintainability, yet they are often time-consuming and cognitively demanding. In response, artificial intelligence (AI)-powered code review tools have gained attention as potential solutions that automate or augment review tasks. This study empirically evaluates the effectiveness of such tools in real-world development environments, with a focus on their capacity to improve defect detection, reduce review time, and support developer productivity. A mixed-methods approach was employed, analysing 25 open-source projects that incorporated both AI-assisted and conventional review practices. Quantitative data were collected on defect detection rates, code turnaround times, and commit frequencies, and statistically compared across projects. To complement this, qualitative insights were gathered from surveys and interviews with developers, exploring trust in AI-generated feedback, perceived benefits, and adoption challenges. The results reveal that AI-powered review tools substantially improve the detection of syntax violations, code smells, and redundant logic, while proving less effective for complex logic, design flaws, and architectural issues that demand human expertise. Productivity gains were also observed, notably in faster review cycles and reduced cognitive load. Nonetheless, concerns were raised regarding false positives, limited contextual understanding, and risks of overreliance on automated feedback. This research contributes empirical evidence to the evolving field of AI in software engineering and provides practical recommendations for integrating AI-powered review tools. Findings underscore that a hybrid approach, blending automation with human judgment, yields the most reliable and effective outcomes.

Harnessing Next-Generation IoT for Sustainable Pharma: Path from Efficient Design to Supply Chain

P. Devi Sravanthi — 2026-01-31

This study looks at how next-generation internet of things (IoT) technologies can improve sustainable practices in developing and optimizing pharmaceutical products. It highlights the shift from eco-design principles to maintaining supply chain integrity. The pharmaceutical industry faces rising demands for sustainability alongside strict regulations and growing consumer expectations. Using IoT advancements creates new chances to improve sustainability throughout the pharmaceutical lifecycle. This study examines how IoT-enabled sensors, data analytics, and connectivity solutions allow real-time monitoring of environmental impacts, resource use, and product quality in the pharmaceutical supply chain. By using IoT capabilities, pharmaceutical companies can take proactive eco-design steps to reduce waste, improve energy use, and lessen their environmental impact during product development and manufacturing. Moreover, IoT-enabled supply chain monitoring provides transparency, traceability, and authenticity. This helps protect against counterfeit drugs and ethical issues. Through case studies, analysis, and insights from the industry, this research paper shows how next-generation IoT technologies can promote sustainable practices and maintain supply chain integrity in the pharmaceutical sector. The study also highlights the need to combine IoT with artificial intelligence and blockchain to improve decision-making, predictive maintenance, and secure data sharing. This complete approach supports global sustainability goals, boosts operational efficiency, and encourages innovation for a greener and more resilient pharmaceutical system.

Proximal Policy Optimisation-driven Stock Trading Using Simple Moving Average Crossover and Higher High-Higher Low Stock Price Structure

Arup Kadia — 2026-02-26

Due to the fluctuations and unpredictability of financial markets, a rule-based trading approach is often ineffective in an environment where conditions are constantly changing. This project is an automated stock trading system that implements the proximal policy optimization (PPO) reinforcement learning approach and signals a simple moving average (SMA) crossover to identify the trend direction and use higher high, higher low (HH, HL) stock price structure to confirm trends with greater certainty. An SMA crossover also identifies those trends that may reverse; meanwhile, an HH-HL structure confirms a bullish price continuation by detecting market microstructures and maintaining the momentum of the market microstructure. The PPO agent is trained on a customised trading environment that models trade execution realities (transaction cost model, position limits, and risk exposure). Consequently, market states are represented as normalised price trends and crossover dynamics and are reflective of structural price action features. The reward function of the agent is designed to maximise risk-adjusted returns while minimising drawdowns and overtrading. The experiment performed by the researcher based on a large sample of historical equity data (multi-year) showed that the proposed method yielded higher cumulative returns, an increased Sharpe ratio, and drawdown reduction compared to conventional SMA strategies and baseline reinforcement learning models. Moreover, the experimental data imply that the use of policy, gradient optimisation techniques combined with trend and structural awareness features, can lead to increased trading stability and adaptability to different market regimes.

SVM-based Medical Diagnosis: A Review of Multicollinearity-aware Feature Selection

Satish Kumar Kalagotla — 2026-04-04

Feature selection is essential in medical diagnosis applications where datasets often contain redundant or highly correlated features. Multicollinearity occurs when predictor variables are strongly correlated, creating challenges for support vector machine (SVM) classifiers, including unstable decision boundaries and reduced generalization performance. This study presents a literature survey of multicollinearity-aware feature selection frameworks for SVM-based medical diagnosis, focusing on correlation-SVM approaches that integrate Pearson correlation analysis and variance inflation factor (VIF) computation. The study examines the theoretical foundations of multicollinearity, its impact on SVM performance, and existing feature selection methodologies, including filter, wrapper, and embedded methods. The survey reveals that correlation-guided feature selection with iterative VIF recomputation consistently achieves 40–63% feature reduction while improving classification accuracy by 2–8% on benchmark medical datasets. The study also explores recent advancements and identifies promising directions for future research, including multi-objective optimization and explainable AI integration.

Artificial Intelligence and Research Literacy among University Students: Opportunities, Risks, and Library Interventions

Md Atique — 2026-04-04

The academic landscape is changing as a result of artificial intelligence (AI), which has a big impact on how university students obtain, assess, and create information. This essay explores the connection between research literacy and AI, stressing the importance of academic libraries in promoting responsible use of AI tools as well as the associated risks and potential. Increased information discovery, individualised learning support, increased research efficiency, and new opportunities for creativity and interdisciplinary inquiry are just a few advantages of AI technologies, such as generative text models, intelligent search systems, and automated summarisation tools. These resources can assist students in better managing their academic obligations and lower obstacles to intricate research procedures. But there are also a lot of difficulties with incorporating AI into academic work. Misinformation and fake content, weakened critical thinking abilities, challenges with academic integrity, ingrained algorithmic bias, and data privacy are among the risks. Students' capacity to critically assess sources and interact extensively with academic materials may be compromised by an over-reliance on AI-generated outputs. As a result, research literacy in the digital age needs to be expanded to include AI literacy, or an awareness of the limitations, ethical implications, and workings of AI systems. According to the paper, university libraries are in a unique position to deal with these issues. Libraries may promote research and AI literacy through focused workshops, curriculum cooperation, AI tool evaluation guides, ethical policy formulation, and the incorporation of critical evaluation frameworks. Universities can guarantee that AI becomes a tool for intellectual empowerment rather than academic dependency by implementing proactive interventions, giving students the tools they need to responsibly and successfully navigate an increasingly AI-mediated research environment.