TASK Quarterly

Analysis of heart rate variability using mobile devices and machine learning

Jan Bancerewicz — 2026-03-26

This paper presents methods of heart rate variability analysis based on electrocardiographic and photoplethysmographic signals obtained using mobile devices. An approach to processing the recorded data involving noise suppression, peak detection, and determining inter-pulse intervals in the physiological signal was developed. Detection of R peaks in ECG and maxima of a PPG wave was performed using convolutional and recurrent neural networks. The accuracy of the models was measured in comparison to the classic algorithms and reference data. The results show a high accuracy of peak identification and allow employing mobile devices to monitor cardiac parameters in both home environments and during physical activities.

A Comparative Analysis of Standard and Graph-Based Retrieval-Augmented Generation

Mikita Kasiak — 2026-03-26

This study investigates the performance of Retrieval-Augmented Generation (RAG) systems, comparing a standard implementation with a graph-driven approach. The paper details the architectural differences between the two systems and presents a comprehensive evaluation of their performance on a diverse dataset. The results demonstrate the advantages of using knowledge graphs to capture relationships between entities, particularly for complex queries. The analysis also considers the trade-offs between performance and resource consumption, providing insights into the practical applications of each approach.

An Analysis of Retrieval-Augmented Generation: A Systematic Review Addressing Architectures, Components, and Evaluation

Adam Ślusarek — 2026-03-26

Retrieval-Augmented Generation (RAG) enhances Large Language Models (LLMs) by integrating external retrieval
mechanisms to improve factuality and currency. This systematic literature review characterizes current RAG architectures, components, and evaluation practices in peer-reviewed studies published between 2021 and 2025 across IEEE
Xplore, Scopus, and Web of Science. Conducted in accordance with the PRISMA guidelines, this review analyzes
41 studies that met the predefined inclusion criteria. Most research addresses Question Answering (QA) and dialogue
systems, employing diverse encoders and retrieval optimization methods. Key findings reveal a strong trend toward integrating OpenAI’s GPT models, alongside growing adoption of open-source alternatives. Persistent challenges include
hallucination control, computational efficiency, and inconsistent evaluation metrics. Despite the potential of RAG, the
evidence base is limited by a focus on English-language, high-resource domains. Furthermore, reproducibility is constrained by heterogeneous evaluation standards and a lack of open-access code or datasets. This review maps the RAG
research landscape and identifies gaps in standardization, scalability, and application to low-resource languages. The
protocol was not prospectively registered, and no funding was received for this work.