10th International Congress on Information and Communication Technology in concurrent with ICT Excellence Awards (ICICT 2025) will be held at London, United Kingdom | February 18 - 21 2025.
Sign up or log in to bookmark your favorites and sync them to your phone or calendar.
Authors - Bruno Zaninotto, Carlos Eduardo Barbosa, Alice Fonseca Monteiro, Lucas Nobrega, Luiz Felipe Martinez, Matheus Argolo, Geraldo Xexeo, Jano Moreira de Souza Abstract - The dynamic between buyers and sellers in the retail sector often leads to conflicts, necessitating a deeper understanding of customer complaints. The Internet is where customers can voice their opinions to influence purchasing decisions and shape company reputations. Brazil, recognized among the top 10 countries with the highest expectations for e-commerce growth worldwide in 2022, demonstrates a rapidly expanding market ready for exploration. This study addresses the problem by applying Latent Semantic Analysis (LSA) to analyze complaints about Americanas S.A., a large retail company on the Reclame Aqui platform, using the company as a case study for broader methodological application. Our findings reveal significant uniformity in complaints across Brazil, primarily concerning order processing, delivery, and product quality. These insights offer actionable intelligence for retailers to refine their Customer Relationship Management strategies and for the government to strengthen consumer protection policies, demonstrating the utility of LSA in improving customer satisfaction and trust in the retail landscape.
Authors - Islam Oshallah, Mohamed Basem, Ali Hamdi, Ammar Mohammed Abstract - Question answering systems face critical limitations in languages with limited resources and scarce data, making the development of robust models especially challenging. The Quranic QA system holds significant importance as it facilitates a deeper understanding of the Quran, a Holy text for over a billion people worldwide. However, these systems face unique challenges, including the linguistic disparity between questions written in Modern Standard Arabic and answers found in Quranic verses written in Classical Arabic, and the small size of existing datasets, which further restricts model performance. To address these challenges, we adopt a cross-language approach by (1) Dataset Augmentation: expanding and enriching the dataset through machine translation to convert Arabic questions into English, paraphrasing questions to create linguistic diversity, and retrieving answers from an English translation of the Quran to align with multilingual training requirements; and (2) Language Model Fine-Tuning: utilizing pre-trained models such as BERT-Medium, RoBERTa-Base, DeBERTa-v3-Base, ELECTRA-Large, Flan-T5, Bloom, and Falcon to address the specific requirements of Quranic QA. Experimental results demonstrate that this cross-language approach significantly improves model performance, with RoBERTa-Base achieving the highest MAP@10 (0.34) and MRR (0.52), while DeBERTa-v3-Base excels in Recall@10 (0.50) and Precision@10 (0.24). These findings underscore the effectiveness of cross-language strategies in overcoming linguistic barriers and advancing Quranic QA systems.
Authors - Aziza Irmatova, Mukhabbatkhon Mirzakarimova, Dilafruz Iskandarova, Guli-ra'no Abdumalikova Abstract - In the today, the development of digital education is playing an important role in radically changing the education system and making learning processes more innovative, interactive and convenient. In particular, digital platforms are the main tools that can change the educational process. Through these platforms, students have the opportunity to study lessons anywhere and at any time, without being limited to traditional classrooms. From this point of view, the development and implementation of digital educational platforms in educational institutions is one of the urgent issues, and the success of this process largely depends on the Internet coverage in the country, investments in digital infrastructure, and the impact of government policy. This article empirically analyzes the impact of Internet coverage, investments in digital infrastructure, and government policy on the implementation of digital educational platforms in Uzbekistan. The measurement of government policy was carried out by assessing the public's assessment of government policy.
Authors - Baraa Hikal, Ahmed Nasreldin, Ali Hamdi, Ammar Mohammed Abstract - Hallucination detection in text generation remains an ongoing struggle for natural language processing (NLP) systems, frequently resulting in unreliable outputs in applications such as machine translation and definition modeling. Existing methods struggle with data scarcity and the limitations of unlabeled datasets, as highlighted by the SHROOM shared task at SemEval-2024. In this work, we propose a novel framework to address these challenges, introducing DeepSeek Few-shot Optimization to enhance weak label generation through iterative prompt engineering. We achieved high-quality annotations that considerably enhanced the performance of downstream models by restructuring data to align with instruct generative models. We further fine-tuned the Mistral-7B-Instruct-v0.3 model on these optimized annotations, enabling it to accurately detect hallucinations in resource-limited settings. Combining this fine-tuned model with ensemble learning strategies, our approach achieved 85.5% accuracy on the test set, setting a new benchmark for the SHROOM task. This study demonstrates the effectiveness of data restructuring, few-shot optimization, and fine-tuning in building scalable and robust hallucination detection frameworks for resource-constrained NLP systems.
Authors - Youssef Maklad, Fares Wael, Wael Elsersy, Ali Hamdi Abstract - This paper presents a novel approach to evaluate the efficiency of a RAG-based agentic Large Language Model (LLM) architecture in network packet seed generation for network protocol fuzzing. Enhanced by chain-of-thought (COT) prompting techniques, the proposed approach focuses on the improvement of the seeds’ structural quality in order to guide protocol fuzzing frameworks through a wide exploration of the protocol state space. Our method leverages RAG and text embeddings in a two-stages. In the first stage, the agent dynamically refers to the Request For Comments (RFC) documents knowledge base for answering queries regarding the protocol Finite State Machine (FSM), then it iteratively reasons through the retrieved knowledge, for output refinement and proper seed placement. In the second stage, we evaluate the response structure quality of the agent’s output, based on metrics as BLEU, ROUGE, and Word Error Rate (WER) by comparing the generated packets against the ground truth packets. Our experiments demonstrate significant improvements of up to 18.19%, 14.81%, and 23.45% in BLEU, ROUGE, and WER, respectively, over baseline models. These results confirm the potential of such approach, improving LLM-based protocol fuzzing frameworks for the identification of hidden vulnerabilities.
Authors - Tushar Vasudev, Surbhi Ranga, Sahil Sankhyan, Praveen Kumar, K V Uday, Varun Dutt Abstract - To guarantee the safety and effectiveness of medical supplies like blood and vaccinations, careful environmental monitoring is necessary throughout transit. Even while real-time monitoring has advanced, current systems sometimes lack strong predictive ability to foresee unfavorable circumstances. The XGBoost Ensemble for Medical Supplies Transport (XEMST), a unique stacking ensemble model created to predict interior humidity levels during travel, is presented in this paper to fill this gap. By utilizing XGBoost's outstanding predictive fusion capabilities, the model incorporates predictions from fundamental machine learning methods, including Support Vector Machine, Random Forest, Decision Tree, and Linear Regression. XEMST outperformed individual models with a Root Mean Squared Error (RMSE) of 2.22% and an R2 score of 0.96 when tested across 17 different transit situations. By enabling prompt responses, these predictive insights protect medical supply quality from environmental hazards. This study demonstrates how sophisticated ensemble learning frameworks have the potential to transform intelligent healthcare logistics.
