Context-Aware Conversational AI for Intelligent Travel Planning and Smart Tourism

2026-09-14T06:29:32+00:00

Artificial Intelligence (AI) has significantly transformed smart tourism by enabling intelligent, personalised, and context-aware travel planning. Recent advancements in Natural Language Processing (NLP), Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), and recommendation systems have enhanced conversational travel assistants by supporting itinerary generation, destination recommendation, and real-time decision-making. This article reviews recent developments in context-aware conversational AI for intelligent travel planning and smart tourism. Existing approaches are analysed based on recommendation techniques, conversational capabilities, contextual awareness, and personalisation. The review also examines key challenges, including data sparsity, limited real-time adaptability, multilingual interaction, privacy and security concerns, and the fragmentation of tourism-related services. Furthermore, emerging technologies such as multimodal AI, agentic AI, and knowledge-enhanced recommendation systems are explored as promising approaches for developing more adaptive and intelligent tourism applications. The study highlights significant research gaps in current conversational travel systems, particularly in maintaining continuous context, integrating diverse real-time data sources, and providing reliable personalised recommendations. Overall, the review emphasizes the potential of context-aware conversational AI to create scalable, adaptive, and user-centric travel-planning systems that can improve decision-making and enhance the overall tourist experience.

Trustworthy AI for Network Intrusion Detection: A Critical Review of Deep Learning, Federated Learning, Explainability, and Large Language Models

2026-10-30T04:57:39+00:00

A network intrusion detector is valuable only when its alerts help defenders make sound decisions. This critical narrative review compares supervised and anomaly-based deep learning, federated detection, explainable AI, and large-language-model (LLM) assistance for network intrusion detection (NIDS). Its principal contribution is a study-level comparison of existing methods and a unified, deployment-oriented assessment of generalization, privacy, adversarial resilience, explanation quality, and analyst workload. The literature synthesis identifies three recurring findings: (i) high scores on static intrusion datasets do not establish cross-site or temporal reliability; (ii) federated data locality does not itself prevent gradient leakage, malicious updates, or unequal performance across clients; and (iii) plausible explanations and LLM-generated incident summaries require evidence tracing, security controls, and human verification. The study recommends chronological and cross-organizational evaluation, operational alert metrics, calibrated thresholds, threat-model-driven testing, and staged deployment validation. The article does not present a newly trained detector or a systematic meta-analysis; its conclusions are qualitative, evidence-qualified recommendations for more trustworthy cybersecurity operations.

International Journal of Computer Science, Algorithms and Programming Languages

Context-Aware Conversational AI for Intelligent Travel Planning and Smart Tourism

Trustworthy AI for Network Intrusion Detection: A Critical Review of Deep Learning, Federated Learning, Explainability, and Large Language Models