Belief, Desire And Intention Research Articles

BDIQA: A New Dataset for Video Question Answering to Explore Cognitive Reasoning through Theory of Mind

As a foundational component of cognitive intelligence, theory of mind (ToM) can make AI more closely resemble human thought processes, thereby enhancing their interaction and collaboration with human. In particular, it can significantly improve a model's comprehension of videos in complex scenes. However, current video question answer (VideoQA) datasets focus on studying causal reasoning within events, few of them genuinely incorporating human ToM. Consequently, there is a lack of development in ToM reasoning tasks within the area of VideoQA. This paper presents BDIQA, the first benchmark to explore the cognitive reasoning capabilities of VideoQA models in the context of ToM. BDIQA is inspired by the cognitive development of children's ToM and addresses the current deficiencies in machine ToM within datasets and tasks. Specifically, it offers tasks at two difficulty levels, assessing Belief, Desire and Intention (BDI) reasoning in both simple and complex scenarios. We conduct evaluations on several mainstream methods of VideoQA and diagnose their capabilities with zero-shot, few-shot and supervised learning. We find that the performance of pre-trained models on cognitive reasoning tasks remains unsatisfactory. To counter this challenge, we undertake thorough analysis and experimentation, ultimately presenting two guidelines to enhance cognitive reasoning derived from ablation analysis.

Trust Model of Privacy-Concerned, Emotionally Aware Agents in a Cooperative Logistics Problem

In this paper, we propose a trust model to be used in a hypothetical mixed environment where humans and unmanned vehicles cooperate. We address the inclusion of emotions inside a trust model in a coherent way to investigate the practical approaches to current psychological theories. The most innovative contribution of this work is the elucidation of how privacy issues play a role in the cooperation decisions of the emotional trust model. Both emotions and trust were cognitively modeled and managed with the beliefs, desires and intentions (BDI) paradigm in autonomous agents implemented in GAML (the programming language of the GAMA agent platform), that communicate using the IEEE FIPA standard. The trusting behavior of these emotional agents was tested in a cooperative logistics problem wherein agents have to move objects to destinations and some of the objects and places are associated with privacy issues. Simulations of the logistic problem show how emotions and trust contribute to improving the performance of agents in terms of both time savings and privacy protection.

Intelligent Software Agent in E-Health System: - Review

The field of medical care is characterized by complexity, dynamism, and variety. Different challenges occur in healthcare (increasing cost of care, population growth and lack of caregivers). For that, intelligent applications with Agent Technology can provide better health care than the conventional medical system. In fact, the characteristics of Intelligent Agents (proactivity, sociability, autonomy) are a good choice to solve many problems in this regard, since the key success parameter it was the natural mapping for real-life medical issues into electronic life. The design of Multi-agent architecture will effectively paradigm the heterogeneous, distributed, and independent health treatment systems. The multi-agent systems have been used of individual healthcare processes such as a knowledge-depend medical system for complex, multi-parties depend on systems such as a total healthcare where The applied from multi-agent systems(mas) in health treatment sector also has been opened avenues for discovering innovative applications such as socialized and personalized health care models. Therefore, when creating an agent-depend simulation, one of the most important questions to consider is how to model the decision decision-making operations for the agents in the model. In the literature, the wide There are a variety of agent decision-making models available, all of them motivated by multiple goals and study questions. The research explains the advances made so far in the development of regulatory frameworks for deliberative autonomous software in a partially observable environment. In this paper, we display an overview for some of an agent that has attracted attention (Beliefs Desires and Intentions (BDI) and its derivatives such as Extensible Beliefs Desires and Intentions (EBDI), Beliefs-Desires-Obligations-Intentions (BOID)).).

Review of Agent Based Modelling of Social Attachment in Crisis Situations

Human behaviour during crisis evacuations is social in nature. In particular, social attachment theory posits that proximity of familiar people, places, objects, etc., promotes calm and a feeling of safety, while their absence triggers panic or flight. In closely bonded groups such as families, members seek each other and evacuate as one. This makes attachment bonds necessary in the development of realistic models of mobility during crises. This article presents a review of evacuation behaviour, theories on social attachment, crisis mobility, and agent-based models. It was found that social attachment influences mobility in the different stages of evacuation (pre, during and post). Based on these findings, a multi-agent model of mobility during seismic crises (SOLACE) is being developed, and it is implemented using the belief, desire and intention (BDI) agent architecture.

Integrated parallel forecasting model based on modified fuzzy time series and SVM

Goals reasoning and management of pilot is a key issue to monitor pilot's behavior and intention. Traditional modeling methods are based on scenarios or situations, such methods will cause the covering problem due to redundancy and are incapable of depicting interactions among various goals and plans of pilot. Petri net integrated with belief, desire and intention (BDI) theory (BDI Petri net) is designed to solve this problem. Focusing on the BDI theory, goal states of agent are discussed firstly. Belief, desire and intention are modeled by places and transitions based on the Petri net theory. In order to simplify the network, colored token is introduced to depict various states of belief, and the hierarchy transition is applied to model the intention, together with tokens' flow demonstrating the interaction among various goals and relationship among belief, desire and intention. A search and rescue mission is used to validate the proposed method and the result indicates that the model can be used to monitor goals and behaviors of pilots.

BDI agents in social simulations: a survey

AbstractModelling and simulation have long been dominated by equation-based approaches, until the recent advent of agent-based approaches. To curb the resulting complexity of models, Axelrod promoted the KISS principle: ‘Keep It Simple, Stupid’. But the community is divided and a new principle appeared: KIDS, ‘Keep It Descriptive, Stupid’. Richer models were thus developed for a variety of phenomena, while agent cognition still tends to be modelled with simple reactive particle-like agents. This is not always appropriate, in particular in the social sciences trying to account for the complexity of human behaviour. One solution is to model humans as belief, desire and intention (BDI) agents, an expressive paradigm using concepts from folk psychology, making it easier for modellers and users to understand the simulation. This paper provides a methodological guide to the use of BDI agents in social simulations, and an overview of existing methodologies and tools for using them.

Fuzzy BDI agents for supply chain monitoring in an uncertain environment

ABSTRACTThe monitoring of supply chains (SCs) is a major challenge due to the increasing complexity of the global market, which increases the exposure of monitoring systems to disruptions, caused by technological innovations and consumer needs. Thus, a dynamic behaviour model to incorporate demand uncertainty in the SC of a Small and Medium Enterprise (SME) was designed in the present study. In our system, information uncertainty is modelled in a fuzzy manner based on Belief, Desire and Intention (BDI) architecture. To achieve our objective, we enhance the multi-agent knowledge model of the SC. Our contribution includes the development of a monitoring system. We first propose a dynamic reasoning model of fuzzy BDI agents faced with an uncertain situation. We then describe the collaborative behaviour between the agents of the SMEs. The challenge addressed in this study is how to make beneficial decisions given uncertain information.

Achieving multi-agent Coo^2 using the BDI & SODA methodologies

The design and implementation of hybrid intelligent systems is gaining pace. Intelligent Multi-Agent System (IMASs) are being designed and developed for dynamic and contested environments, allowing systems to exercise and act upon unfamiliar, and at times progressive events. Hybrid and intelligent systems require an I-MAS to support multiple participants, each with its own view of the environment and that has the ability to perceive, decide and act on events. To ensure cohesion and goal-driven behaviours are conducted, Coordination and Cooperation (Coo) techniques need to be revisited. Current Coo techniques fail to recognise the importance of mental modelling and situation assessment of agents. Moreover, cognitive and reasoning architectures, like Observe-Orient-Decide-Act (OODA) and Belief, Desire and Intention (BDI) do not consider how organisational principles, like Coo, can assist in how an agent behaves when dealing with an event. The Agent Coordination and Cooperation Cognitive Model (ACM) is a new generation of agent-based models that will exploit an agent’s cognitive and mental architectures to strengthen Coo. This is achieved by establishing processes that exhibit how an agent’s stimulation, orientation and belief-set is related to its ability to coordinate and be coordinated. ACM also illustrates how decision-making processes and an agent’s ability to act is connected to its desires and intentions. In addition, the relationship between decision and action and desires and intentions provides the foundations for cooperation between agents. Integrating elements within BDI and Stimulate-Orient-Decide-Act (SODA) produces superior and complex Coo techniques. Consequently, by associating beliefs and orientation to coordination, as well as desire, intentions to cooperation will strengthen and improve situation awareness and decision-making, which is critical in dynamic environments.

The integration of cognitive architectures and computational intelligence using the SODA methodology

The versatility and application of Intelligent Multi-Agent Systems (I-MASs) is becoming more complex. The increase in complexity is due to a sudden requirement for Intelligent Agents (IAs) and I-MASs to exhibit more human-like functionality. Thisincludes the abilityto represent knowledge using perception, learning, recognition, reasoning and rationality. In addition, this wishlist of functionality also needs to be constrained within the limits of current technology. Designers of I-MASs are looking at mechanisms and processes that can satisfy more involved requirements. Two such paradigms are cognitive architectures and computational intelligence. The Agent Coordination and Cooperation Cognitive Model (AC 3 M) was developed to primarily illustrate the link between agent coordination and cooperation. Whilst analysing AC 3 M, the Stimulate-Orient-Decide-Act (SODA) methodology was discovered. This methodology is a result of integrating cognitive architecture principles with computational intelligence. SODA was the integration of the Belief, Desire and Intention (BDI) methodology with the Observe-Orient-Decide-Act (OODA) loop and results in the components of BDI being linked to phases of the OODA loop. The main advantage of the SODA methodology is the enhancement of an agent's situational awareness, which has proven to be paramount in the area of team automation.

Coalition utility allocation based on agent's outstanding characteristics

Coalition formation is a key topic in multi-agent system (MAS), considering the existing strategies cannot clearly distinguish each agent's contribution in MAS, which may result in coalition's potential instability as well as the low timeliness. In order to tackle the shortage above, this paper, based on belief desire and intention (BDI) model, proposes a novel coalition utility allocation strategy through updating agents' outstanding characteristics. According to different requirements of task, agents are eager to update their outstanding characteristics timely in order to maximise their own benefit and ensure friendly cooperation with other members to finish the task as well. The theoretical analysis and experiments show that by the novel strategy, the utility is reasonably distributed to each agent, what's more, it could improve the agents' adaptation to environment and basically satisfy all kinds of requirements in super-additive task oriented domains, such as the reasonable allocation of benefit among agents; the global optimal solution, which is strong stability, timeliness and distributed.

A first-order coalition logic for BDI-agents

Coalition logic (CL) enables us to model the strategic abilities and specify what a group of agents can achieve whatever the other agents do. However, some rational mental attitudes of the agents are beyond the scope of CL such as the prestigious beliefs, desires and intentions (BDI) which is an interesting and useful epistemic notion and has spawned substantial amount of studies in multi-agent systems. In this paper, we introduce a first-order coalition BDI (FCBDI) logic for multi-agent systems, which provides a semantic glue that allows the formal embedding and interaction of BDI, coalition and temporal operators in a first-order language. We further introduce a semantic model based on the interpreted system model and present an axiomatic system that is proved sound and complete with respect to the semantics. Finally, it is shown that the computational complexity of its model checking in finite structures is PSPACE-complete.

Integrated control of lateral and vertical vehicle dynamics based on multi-agent system

The existing research of the integrated chassis control mainly focuses on the different evaluation indexes and control strategy. Among the different evaluation indexes, the comprehensive properties are usually not considered based on the non-linear superposition principle. But, the control strategy has some shortages on tyre model with side-slip angle, road adhesion coefficient, vertical load and velocity. In this paper, based on belief, desire and intention(BDI)-agent model framework, the TYRE agent, electric power steering(EPS) agent and active suspension system(ASS) agent are proposed. In the system(SYS) agent, the coordination mechanism is employed to manage interdependences and conflicts among other agents, so as to improve the flexibility, adaptability, and robustness of the global control system. Due to the existence of the simulation demand of dynamic performance, the vehicle multi-body dynamics model is established by SIMPACK. And then the co-simulation analysis is conducted to evaluate the proposed multi-agent system(MAS) controller. The simulation results demonstrate that the MAS has good effect on the performance of EPS and ASS. Meantime, the better road feeling for the driver is provided considering the multiple and complex driving traffic. Finally, the MAS rapid control prototyping is built to conduct the real vehicle test. The test results are consistent to the simulation results, which verifies the correctness of simulation. The proposed research ensures the driving safety, enhances the handling stability, and improves the ride comfort.

A Multi-Agent-Based Intelligent Sensor and Actuator Network Design for Smart House and Home Automation

The smart-house technology aims to increase home automation and security with reduced energy consumption. A smart house consists of various intelligent sensors and actuators operating on different platforms with conflicting objectives. This paper proposes a multi-agent system (MAS) design framework to achieve smart house automation. The novelties of this work include the developments of (1) belief, desire and intention (BDI) agent behavior models; (2) a regulation policy-based multi-agent collaboration mechanism; and (3) a set of metrics for MAS performance evaluation. Simulations of case studies are performed using the Java Agent Development Environment (JADE) to demonstrate the advantages of the proposed method.

Levels of modality for BDI Logic

The use of rational agents for modelling real world problems has both been heavily investigated and become well accepted, with BDI (Beliefs, Desires, and Intentions) Logic being a widely used architecture to represent and reason about rational agency. However, in the real world, we often have to deal with different levels of confidence in the beliefs we hold, desires we have, and intentions that we commit to. This paper extends our previous framework that integrated qualitative levels of beliefs, desires, and intentions into BDI Logic. We describe an expanded set of axioms and properties of the extended logic. We present a modular structure for the semantics which involves a non-normal Kripke type semantics that may be used for other agent systems. Further, we demonstrate the usefulness of our framework with a scheduling task example.

Cognitive Agent Based Identification of Relevant Auctions in Mobile E-commerce

The increased attention of E-auction services in mobile E-commerce demands an approach to identify the relevant auctions as per the bidder’s requirements so as to increase the bidder satisfaction level and auction winning probability. In this paper, we propose an intelligent agent based model to identify the relevant set of auctions for a mobile bidder based on the bidder’s requirements, preferences and constraints from a set of active auctions available in the active auction service directory in regional gateway connected to Internet. The agent functions are based on Belief, Desire and Intention (BDI) cognitive architecture and are capable of taking dynamic decisions to search the matching auctions for the products requested by the bidder in the bidder belief set located in the bidder’s mobile device and/or in the case base of the regional gateway. If matching auctions are not found in either of them, BDI agent searches them in the active auction service directory and computes the relevance factor based on the parameters in the bidder’s requirements for all the matching active auctions and clusters into relevant (or potential) and non relevant auctions. The model is simulated to test the performance measures like availability of relevant auctions, average response time, and probability of winning auctions with better satisfaction. The proposed model is also compared with advertisement based auction service discovery model to show its effectiveness.

REPLANNING MECHANISM FOR DELIBERATIVE AGENTS IN DYNAMIC CHANGING ENVIRONMENTS

This paper proposes a replanning mechanism for deliberative agents as a new approach to tackling the frame problem. We propose a beliefs desires and intentions (BDI) agent architecture using a case‐based planning (CBP) mechanism for reasoning. We discuss the characteristics of the problems faced with planning where constraint satisfaction problems (CSP) resources are limited and formulate, through variation techniques, a reasoning model agent to resolve them. The design of the agent proposed, named MRP‐Ag (most‐replanable agent), will be evaluated in different environments using a series of simulation experiments, comparing it with others such as E‐Ag (Efficient Agent) and O‐Ag (Optimum Agent). Last, the most important results will be summarized, and the notion of an adaptable agent will be introduced.

A Role-Based Approach in Dynamic Task Delegation in Agent Teamwork

This paper presents a role-based agent teamwork language called RoB-MALLET (Role-Based Multi-Agent Logic Language for Encoding Teamwork). Roles have been used to form multi-agent theories and guide the construction of multi-agent architectures and systems. Different from the most existing agent teamwork models that emphasize practical reasoning based on beliefs, desires and intentions (BDI), a role-based teamwork language stresses the concept of organization and thereby cooperation. RoB-MALLET is able to handle unexpected uncertainties and allow for the conceptual specification of teamwork knowledge for reuse. Finally, we present several experiments that explore plan reusability of RoB-MALLET, show that it is flexible in supporting simultaneity of invocation and evaluate its efficiency in supporting teamwork.

Intention without representation

A mechanism for planning ahead would appear to be essential to any creature with more than insect level intelligence. In this paper it is shown how planning, using full means‐ends analysis, can be had while avoiding the so called symbol grounding problem. The key role of knowledge representation in intelligence has been acknowledged since at least the enlightenment, but the advent of the computer has made it possible to explore the limits of alternate schemes, and to explore the nature of our everyday understanding of the world around us. In particular, artificial intelligence (AI) and robotics has forced a close examination, by people other than philosophers, of what it means to say for instance that “snow is white.” One interpretation of the “new AI” is that it is questioning the need for representation altogether. Brooks and others have shown how a range of intelligent behaviors can be had without representation, and this paper goes one step further showing how intending to do things can be achieved without symbolic representation. The paper gives a concrete example of a mechanism in terms of robots that play soccer. It describes a belief, desire and intention (BDI) architecture that plans in terms of activities. The result is a situated agent that plans to do things with no more ontological commitment than the reactive systems Brooks described in his seminal paper, “Intelligence without Representation.”

Interrelations between the BDI primitives: Towards heterogeneous agents

The study of formal theories of agents has intensified over the last decade since such formalisms can be viewed as providing the specifications for building agent-based systems. One such theory views agents as having beliefs, desires and intentions (BDI). The BDI paradigm provides us with the means of describing different types of agents; a desirable quality, since agent-based systems are employed in various domains with diverse characteristics and therefore different requirements. This is accomplished by adopting a set of constraints that describe how the three attitudes are related to each other, called a notion of realism. Although three such notions have been explored in the literature, namely strong realism, realism and weak realism, no systematic attempt has been undertaken to study other available options. In this paper we explore the dynamics and possible interrelations between the three attitudes and we propose notions of realism for heterogeneous BDI agents. We explore a wider range of possibilities by considering a combination of the types of relations between accessible worlds. Moreover, we distinguish between two broad categories of agents, bold and circumspect, according to the relation between beliefs and intentions. We explore several interesting notions of realism for such agents and we argue that these come close to the desiderata for rational BDI agents.

Gulliver's Genie: a multi-agent system for ubiquitous and intelligent content delivery

This paper introduces Gulliver's Genie a context-aware tourist guide that assists roaming tourists. The approach adopted within this system is the deployment of intelligent agents, which collectively determine the user context and retrieve and assemble multi-media presentations that are wirelessly transmitted and displayed on a Personal Digital Assistant (PDA). As a backdrop, we first consider the state of the art in terms of context sensitive tourist guides, telecommunications, positioning technology and agent technologies. Gulliver's Genie considers user context in terms of position, orientation and user profile. System agents are strong intentional agents that base deductions on a mental state comprising of Beliefs, Desires and Intentions (BDI). This paper presents the design of the system together with a glimpse of the user experience.