https://revistas.usal.es/cinco/index.php/2255-2863 <p dir="ltr">The <a title="adcaij" href="http://adcaij.usal.es" target="_blank" rel="noopener">Advances in Distributed Computing and Artificial Intelligence Journal</a> (ISSN: 2255-2863) is an open access (OA) journal that publishes articles which contribute new results associated with distributed computing and artificial intelligence, and their application in different areas, such as the Deep Learning, Generative AI, Electronic commerce, Smart Grids, IoT, Distributed Computing and so on. These technologies are changing constantly as a result of the large research and technical effort being undertaken in both universities and businesses. Authors are solicited to contribute to the journal by submitting articles that illustrate research results, projects, surveying works and industrial experiences that describe significant advances in the areas of computing.</p> <p dir="ltr">ADCAIJ focuses attention in the exchange of ideas between scientists and technicians. Both, academic and business areas, are essential to facilitate the development of systems that meet the demands of today's society. The journal is supported by the research group <a title="bisite" href="http://bisite.usal.es/en/research/research-lines" target="_blank" rel="noopener">BISITE</a>.</p> <p dir="ltr">The journal commenced publication in 2012 with quarterly periodicity and has published more than 300 articles with peer review. All the articles are written in scientific English language.</p> <p dir="ltr">From volume 12 (2023) onwards, the journal will be published in continuous mode, in order to advance the visibility and dissemination of scientific knowledge.</p> <p dir="ltr">ADCAIJ is indexed in Scopus and in the Emerging Sources Citation Index (ESCI) of Web of Science, in the category COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE. It also appears in other directories and databases such as DOAJ, ProQuest, Scholar, WorldCat, Dialnet, Sherpa ROMEO, Dulcinea, UlrichWeb, BASE, Academic Journals Database and Google Scholar.</p> Universidad de Salamanca en-US ADCAIJ: Advances in Distributed Computing and Artificial Intelligence Journal 2255-2863 https://revistas.usal.es/cinco/index.php/2255-2863/article/view/31782 Prefetching is essential to minimizing the number of misses in cache and improving processor performance. Many prefetchers have been proposed, including simple but highly effective stream-based prefetchers and prefetchers that predict complex access patterns based on structures such as history buffers and bit vectors. However, many cache misses still occur in many applications. After analyzing the various techniques in Instruction and Data Prefetcher, several key features were extracted which impact system performance. Data prefetching is an essential technique used in all commercial processors. Data prefetchers aim at hiding the long data access latency. In this paper, we present the design of an Adaptive Delta-based Data Prefetching (ADDP) that employs four different tables organized in a hierarchical manner to address the diversity of access patterns. Firstly, the Entry Table is queue, which tracks recent cache fill. Secondly, the Predict Table which has trigger (Program Counter) PCs as tags. Thirdly, the (Address Difference Table) ADT which has target PCs as tags. Lastly, the Prefetch Table is divided into two parts, i.e., Prefetch Filter and the actual Prefetch Table. The Prefetch Filter table filters unnecessary prefetch accesses and the Prefetch Table is used to track other additional information for each prefetch. The ADDP has been implemented in a multicache-level prefetching system under the 3rd Data Prefetching Championship (DPC-3) framework. ADDP is an effective solution for data-intensive applications since it shows notable gains in cache hit rates and latency reduction. The simulation results show that ADDP outperforms the top three data prefetchers MLOP, SPP and BINGO by 5.312 %, 13.213 % and 10.549 %, respectively. Swapnita Srivastava P. K. Singh Copyright (c) 2025 Swapnita Srivastava , P K Singh https://creativecommons.org/licenses/by-nc-nd/4.0/ 2025-04-11 2025-04-11 14 e31782 e31782 10.14201/adcaij.31782 https://revistas.usal.es/cinco/index.php/2255-2863/article/view/31857 This comparative analysis delves into the capabilities of three prominent Conversational AI models: ChatGPT, Google Bard, and Microsoft Bing Chat. The study encompasses a meticulous exploration of their conversational skills, natural language processing abilities, and creative text generation. Methodologically, this study crafts a comprehensive evaluation framework, including complexity levels and tasks for each dimension. Through user-generated responses, key metrics of fluency, coherence, relevance, accuracy, completeness, informativeness, creativity, and relevance were assessed. The results reveal distinctive strengths of the AI models. The theoretical implications lead to recommendations for dynamic learning, ethical considerations, and cross-cultural adaptability. Practically, avenues for future research were proposed, including real-time user feedback integration, multimodal capabilities exploration, and collaborative human-AI interaction studies. The analysis sets the stage for benchmarking and environmental impact assessments, underlining the need for standardized metrics. Pascal Adomako Talha Ali Khan Raja Hashim Ali Rand Koutaly Copyright (c) 2025 pascal Adomako, Rand Kouatly, Talha Ali Khan https://creativecommons.org/licenses/by-nc-nd/4.0/ 2025-04-11 2025-04-11 14 e31857 e31857 10.14201/adcaij.31857 https://revistas.usal.es/cinco/index.php/2255-2863/article/view/31737 Diabetes, hypertension, obesity, glaucoma, etc. are severe and common retinopathy diseases today. Early age detection and diagnosis of these diseases can save human beings from many life threats. The retina’s blood vessels carry details of retinopathy diseases. Therefore, feature extraction from blood vessels is essential to classify these diseases. A segmented retinal image is only a vascular tree of blood vessels. Feature extraction is easy and efficient from segmented images. Today, there are existing different approaches in this field that use RGB images only to classify these diseases due to which their performance is relatively low. In the work, we have proposed a model based on machine learning that uses segmented retinal images generated by different efficient methods to classify diabetic retinopathy, glaucoma, and multi-class diseases. We have carried out extensive experiments on numerous images of DRIVE, HRF, STARE, and RIM-ONE DL datasets. The highest accuracy of the proposed approach is 90.90 %, 95.00 %, and 92.90 % for diabetic retinopathy, glaucoma, and multi-class diseases, respectively, which the model detected better than most of the methods in this field. Sushil Kumar Saroj Rakesh Kumar Nagendra Pratap Singh Copyright (c) 2024 Sushil Kumar Saroj https://creativecommons.org/licenses/by-nc-nd/4.0/ 2025-02-27 2025-02-27 14 e31737 e31737 10.14201/adcaij.31737 https://revistas.usal.es/cinco/index.php/2255-2863/article/view/31698 The current computing era enables the generation of vast amounts of data, which must be processed to extract valuable insights. This processing often requires multiple query operations, where hashing plays a crucial role in accelerating query response times. Among hashing techniques, Cuckoo Hashing has demonstrated greater efficiency than conventional methods, offering simplicity and ease of integration into various real-world applications. However, Cuckoo Hashing also has limitations, including data collisions, data loss due to collisions, and the potential for endless loops that lead to high insertion latency and frequent rehashing. To address these challenges, this work introduces a modified Matrix hashing technique. The core concept of the proposed scheme is to utilize both a 2D array and an additional 1D array with random probing to create a more robust technique that competes effectively with Cuckoo Hashing. This study also introduces degree of dexterity as a new performance metric, in addition to the traditional load factor. Furthermore, the Even-Odd hash function is proposed to ensure a more balanced load distribution. Through rigorous experimental analysis in a single-threaded environment, this modified Matrix hashing with random probing in the 1D array is shown to effectively resolve key issues associated with Cuckoo Hashing, such as excessive data migration, inefficient memory usage, and high insertion latency. Rajeev Ranjan Kumar Tripathi Pradeep Kumar Singh Sarv Pal Singh Copyright (c) 2024 Rajeev Ranjan, Prof.P.K.Singh, Prof.S.P.Singh https://creativecommons.org/licenses/by-nc-nd/4.0/ 2025-02-27 2025-02-27 14 e31698 e31698 10.14201/adcaij.31698 https://revistas.usal.es/cinco/index.php/2255-2863/article/view/31736 In an era where digital transformation is accelerating rapidly, artificial intelligence and machine learning have emerged as transformative forces, especially in e-commerce. This paper presents a comprehensive literature review that delves into the fundamentals of e-commerce, artificial intelligence, and machine learning, highlighting their key advantages and practical applications. By examining a broad array of studies, this research evaluates the critical role of artificial intelligence and machine learning in reshaping e-commerce and explores the potential these technologies hold for enhancing customer engagement and driving sales. The paper underscores how e-commerce companies leverage artificial intelligence-driven innovations to influence customer behaviour, enhance personalised marketing, and streamline purchasing pathways. However, the path to successful artificial intelligence integration is not without obstacles. Challenges such as organisational resistance, skills shortages, technical limitations, and awareness gaps are notable barriers. Despite these hurdles, the findings suggest that adopting artificial intelligence and machine learning tools positions e-commerce companies for long-term success, offering significant competitive advantages and fostering sustainable growth in an increasingly digital world. Fedorko Richard Kráľ Štefan Kráľová Lenka Copyright (c) 2025 Richard Fedorko, Štefan Kráľ, Lenka Štofejová https://creativecommons.org/licenses/by-nc-nd/4.0/ 2025-02-27 2025-02-27 14 e31736 e31736 10.14201/adcaij.31736 https://revistas.usal.es/cinco/index.php/2255-2863/article/view/31609 Blockchain technology has experienced significant growth across various industries. However, challenges such as scalability, high transaction fees, and resource inefficiencies continue to limit its full potential. This paper presents a novel approach using a multilayer graph to model and analyze blockchain transactions, with a focus on resource consumption—specifically opcode execution and gas usage. By categorizing accounts into distinct layers—Externally Owned Accounts (EOAs), smart contracts, oracles, and cross-chain bridges—the graph-based model captures interactions across these account types. Through transaction trace analysis, we extract opcode usage and gas consumption, applying graph-theoretical metrics such as node scoring and edge weighting to identify critical nodes and resource-intensive transactions. Our findings provide new insights into resource-heavy behaviors, revealing optimization opportunities to reduce transaction costs and improve scalability. Additionally, the approach aids in anomaly detection and smart contract optimization, enhancing the cost-effectiveness and performance of blockchain systems. Inas Hasnaoui Maria Zrikem Raja Elassali Copyright (c) 2025 Inas Hasnaoui, Maria Zrikem, Rajaa Elassali https://creativecommons.org/licenses/by-nc-nd/4.0/ 2025-02-27 2025-02-27 14 e31609 e31609 10.14201/adcaij.31609