The rapid expansion of the Internet of Things (IoT) has introduced significant cybersecurity challenges across healthcare, industrial systems, and smart cities. Existing security models provide limited protection due to weak authentication, heterogeneous device architectures, and constrained computational resources. Although Artificial Intelligence (AI) methods, particularly Long Short-Term Memory (LSTM) networks for anomaly detection and Reinforcement Learning (RL) for adaptive defense, show strong potential, most studies emphasize performance metrics while overlooking scalability, energy consumption, and deployability. Concurrently, classical cryptography approaches such as Rivest–Shamir–Adleman (RSA) and Elliptic Curve Cryptography (ECC) are increasingly vulnerable to quantum computing, creating an urgent need for Post-Quantum Cryptography (PQC). This study addresses these gaps by exploring how AI techniques and PQC can jointly strengthen IoT systems in a manner that is adaptive, resource-efficient, and resistant to quantum-level attacks. A systematic literature review (SLR) was conducted following PRISMA 2020 guidelines and Kitchenham’s software engineering methodology. The review examined 150 peer-reviewed articles published between 2014 and 2025 from IEEE, ACM, Springer, ScienceDirect, Scopus, and Web of Science. Findings show that LSTM models consistently achieve high anomaly detection accuracy (91–99%) but impose computational burdens that limit real-time scalability. RL-based defense mechanisms provide strong adaptability for threat response, yet require careful balancing of resource efficiency and response flexibility and currently lack standardization. PQC methods, particularly lattice-based schemes such as Kyber, LWE, and NTRU, demonstrate promise for securing constrained IoT environments against quantum threats, but introduce challenges related to key sizes, hardware demands, and deployment feasibility. Integrated AI-PQC models remain in early development, with limited practical implementations and several unresolved trade-offs. Key challenges include aligning AI’s computational requirements with PQC’s increased resource consumption, ensuring interoperability under strict hardware constraints, and minimizing system-wide latency while maintaining reliability. This review establishes an initial roadmap for developing practical AI-PQC security solutions for IoT, highlighting essential trade-offs among security strength.
Background This article defines an artificial intelligence framework to detect individual’s mental health (MH) status on social networks. The proposed framework, which consists of four main modules, aims to analyze the emotions that are expressed by social network users in their text posts and identify their mental coping strategies, resources, and demands based on The Demands-Resources-Individual Effects (DRIVE) model. Although sentiment analysis (SA) is effective in analyzing the polarity of the text, it is limited in detecting the mental health status in terms of the coping strategies, available resources, or encountered stressors. This study illustrates such limitations in detecting the coping strategies and shows the effectiveness of the coping-based analysis. The work also reveals the phrases and topics that were used by individuals to express their coping strategies which provides a novel outlook of the individuals’ psychological coping within their environment. Methods The social network X is used to collect the coping strategies expressed by people who experienced stress during COVID-19 from November 2019 to May 2022. Text was processed using natural language processing (NLP). A sample of posts was coded into a positive or negative coping category and one of eight subtypes. SA and statistical analysis were performed to compare SA results with coded coping strategies. Latent Dirichlet Allocation and bigram NLP were applied to identify main themes and terminologies. Coping classification models were created and tested. Results The findings reveal that 70% of posts show positive coping strategies. The main positive coping themes included self-care, seeking help, positive reframing, engaging in prayers and meditation, employing humor through sarcasm, and implementing a practical mindset. Conversely, the remaining 30% of posts expressed negative coping themes, such as conspiracy thoughts, wishful or hopeless thinking, and negative perceptions. The coping classification models achieved a reliable predictive level with an average accuracy of 74.8%. Categorizing coping strategies using SA methods, particularly TextBlob and VADER, revealed high miscategorization rates, especially for negative coping strategies. Bigrams and LDA analysis identified distinct word patterns in positive and negative coping strategies, with emojis playing a significant role in emotional expression across both categories. Conclusion The article defined a framework for a MH detector based on the DRIVE model. It highlighted the resilience and adaptive responses of individuals in times of crisis. It also focused on coping and identified physical, emotional, and social support and positive reframing as major positive strategies; and the spread of false information and loss of social support as negative coping strategies. The applied coping classification models showed reliable performance in distinguishing between positive and negative coping categories.
The concepts used in IFOL have associated to them a list of sorted attributes, and the sorts are the intensional concepts as well. The requirement to extend the unsorted IFOL (Intensional FOL) to many-sorted IFOL is mainly based on the fact that a natural language is implicitly many-sorted and that we intend to use IFOL to support applications that use natural languages. Thus, the proposed version of many-sorted IFOL is just the completion of this conceptual feature of the IFOL.
This paper aims to explore the separation of the two forward passes in the Forward-Forward algorithm from a biological perspective in the context of sleep. We show the size of the gap between the sleep and awake phase influences the learning capabilities of the algorithm and highlight the importance of negative data in diminishing the devastating effects of sleep deprivation.
A series of recent papers raises worries about the manipulative potential of algorithmic transparency. But while the concern is apt and relevant, it is based on a fraught understanding of manipulation. Therefore, this paper draws attention to the indifference view of manipulation, which explains better than the vulnerability view why algorithmic transparency has manipulative potential. The paper also raises pertinent research questions for future studies of manipulation in the context of algorithmic transparency.
The paper points at the grieving problems implied by the richness axiom in the Kleinberg's axiomatic system and suggests resolutions. The richness induces learnability problem in general and leads to conflicts with consistency axiom. As a resolution, learnability constraints and usage of centric consistency or restriction of the domain of considered clusterings to super-ball-clusterings is proposed.
We present argumentation schemes to model reasoning with legal cases. We provide schemes for each of the three stages that take place after the facts are established: factor ascription, issue resolution and outcome determination. The schemes are illustrated with examples from a specific legal domain, US Trade Secrets law, and the wider applicability of these schemes is discussed.
Current measures of machine intelligence are either difficult to evaluate or lack the ability to test a robot's problem-solving capacity in open worlds. We propose a novel evaluation framework based on the formal notion of MacGyver Test which provides a practical way for assessing the resilience and resourcefulness of artificial agents.
Representing knowledge as high-dimensional vectors in a continuous semantic vector space can help overcome the brittleness and incompleteness of traditional knowledge bases. We present a method for performing deductive reasoning directly in such a vector space, combining analogy, association, and deduction in a straightforward way at each step in a chain of reasoning, drawing on knowledge from diverse sources and ontologies.
In this short note we address the issue of expressing norms (such as obligations and prohibitions) in temporal logic. In particular, we address the argument from [Governatori 2015] that norms cannot be expressed in Linear Time Temporal Logic (LTL).
This document provides the foundations behind the functionality provided by the $ρ$G library (https://github.com/santiontanon/RHOG), focusing on the basic operations the library provides: subsumption, refinement of directed labeled graphs, and distance/similarity assessment between directed labeled graphs. $ρ$G development was initially supported by the National Science Foundation, by the EAGER grant IIS-1551338.
We describe a framework for building abstraction hierarchies whereby an agent alternates skill- and representation-acquisition phases to construct a sequence of increasingly abstract Markov decision processes. Our formulation builds on recent results showing that the appropriate abstract representation of a problem is specified by the agent's skills. We describe how such a hierarchy can be used for fast planning, and illustrate the construction of an appropriate hierarchy for the Taxi domain.
This paper discusses an expert system shell that integrates rule-based reasoning and the Dempster-Shafer evidence combination scheme. Domain knowledge is stored as rules with associated belief functions. The reasoning component uses a combination of forward and backward inferencing mechanisms to allow interaction with users in a mixed-initiative format.
This paper introduces the notion of objection-based causal networks which resemble probabilistic causal networks except that they are quantified using objections. An objection is a logical sentence and denotes a condition under which a, causal dependency does not exist. Objection-based causal networks enjoy almost all the properties that make probabilistic causal networks popular, with the added advantage that objections are, arguably more intuitive than probabilities.
Irrelevance-based partial MAPs are useful constructs for domain-independent explanation using belief networks. We look at two definitions for such partial MAPs, and prove important properties that are useful in designing algorithms for computing them effectively. We make use of these properties in modifying our standard MAP best-first algorithm, so as to handle irrelevance-based partial MAPs.
Incidence Calculus and Dempster-Shafer Theory of Evidence are both theories to describe agents' degrees of belief in propositions, thus being appropriate to represent uncertainty in reasoning systems. This paper presents a straightforward equivalence proof between some special cases of these theories.
Nonmonotonic reasoning is a pattern of reasoning that allows an agent to make and retract (tentative) conclusions from inconclusive evidence. This paper gives a possible-worlds interpretation of the nonmonotonic reasoning problem based on standard decision theory and the emerging probability logic. The system's central principle is that a tentative conclusion is a decision to make a bet, not an assertion of fact. The system is rational, and as sound as the proof theory of its underlying probability log.