Abstract This study examines the similarity, credibility, and value of information content generated by Google and various Generative AI (GAI) platforms, such as ChatGPT, Qwen, Perplexity, and Poe. Using the Jaccard Similarity Index, the study measures content overlap... This study examines the similarity, credibility, and value of information content generated by Google and various Generative AI (GAI) platforms, such as ChatGPT, Qwen, Perplexity, and Poe. Using the Jaccard Similarity Index, the study measures content overlap between Google search results and AI-generated responses, resulting in two clusters. The first one is the low correlation between the Google and AI outputs. Second is the consistency among AI-based models in terms of similarity. Thus, AI platforms reflect higher internal similarity, suggesting a high degree of homogeneity among them. The research also evaluates credibility, finding Google to have higher average credibility scores (0.82) compared to Perplexity (0.59) in reference analysis, with GAI platforms often lacking reliable citations and traceability. Content credibility scores, assessed using criteria such as source credibility, content accuracy, completeness, objectivity, and verifiability, indicate that Google excels the AI platforms. Google scored 6.54, whereas Perplexity, Qwen, ChatGPT, and Poe scored 4.72, 3.75, 2.81, and 2.26, respectively, in user evaluation. In terms of credibility, Google surpasses generative AI platforms; while AI tools provide useful summaries, they lack the depth, evidence base, and scholarly context found in Google search results. The study concludes that while AI enhances communication efficiency, it raises concerns about reliability, knowledge foraging, and scientific integrity. Future research aims to analyze content volume and length as measures of value, emphasizing the need for transparency and human-centered AI development. Read More Read Less

Abstract Assistance with the use of technical devices is required as soon as the tasks become complex. This assistance is also needed as soon as we provide users with solutions to incidents that occur during the application... Assistance with the use of technical devices is required as soon as the tasks become complex. This assistance is also needed as soon as we provide users with solutions to incidents that occur during the application of unsuitable procedures. The goal of this work is then to provide a knowledge extraction approach that can interpret and identify user requests as valid system requests, thereby responding appropriately to novice user requests. This approach is based on a fuzzy semantic network for modelling imprecise and uncertain knowledge and the automatic construction of a temporary fuzzy ontology for identifying and interpreting user requests. The proposed approach has the advantage of integrating the notion of uncertain and imprecise knowledge into the representation of system objects and procedures. The experimental results show the feasibility, efficiency and effectiveness of our approach. Read More Read Less

Abstract This brief review examines adaptive and enhanced Retrieval-Augmented Generation (RAG) systems, focusing on overcoming the limitations of standard RAG models, such as inefficiency, excessive resource utilisation, and rigid retrieve-then-generate workflows. It highlights advancements like Dynamic RAG... This brief review examines adaptive and enhanced Retrieval-Augmented Generation (RAG) systems, focusing on overcoming the limitations of standard RAG models, such as inefficiency, excessive resource utilisation, and rigid retrieve-then-generate workflows. It highlights advancements like Dynamic RAG and Parametric RAG, which enable context-aware retrieval and parameter-level knowledge integration. The paper emphasises adaptive mechanisms that intelligently decide when and how much to retrieve, improving efficiency and relevance. It also discusses query rewriting, verification, and multimodal extensions to enhance precision. Furthermore, the integration of forecasting latent retrieval methods is introduced, where deep latent dynamics models extract predictable, interpretable components from limited time series data, improving forecasting accuracy. Applications in education, edge computing, and domain-specific contexts are explored, showing reduced hallucinations and better scalability. The review outlines the evolution of RAG toward intelligent, responsive systems using reinforcement learning and hybrid architectures. We identified several potential future research directions, including agentic workflows, multi-modal adaptation, and domainspecific RAG models, which promise more reliable, scalable, and context-aware generative AI applications across various fields. Read More Read Less