Domain-Specific Multi-Agent Dialog Policy Learning in Multi-Domain Task-Oriented Scenarios

2021 ◽  
Author(s):  
Li Tang ◽  
Yuke Si ◽  
Longbiao Wang ◽  
Jianwu Dang
Keyword(s):  
Policy Learning ◽  
Domain Specific ◽  
Multi Agent ◽  
Task Oriented

scholarly journals Multi-Agent Task-Oriented Dialog Policy Learning with Role-Aware Reward Decomposition

2020 ◽  
Author(s):  
Ryuichi Takanobu ◽  
Runze Liang ◽  
Minlie Huang
Keyword(s):  
Policy Learning ◽  
Multi Agent ◽  
Task Oriented

Modular Representations of Cognitive Phenomena in AI, Psychology and Neuroscience

2011 ◽  
pp. 66-89 ◽  
Author(s):  
Joanna J. Bryson
Keyword(s):  
Artificial Intelligence ◽  
Cognitive Science ◽  
Conscious Experience ◽  
Modular Representations ◽  
Domain Specific ◽  
Massive Modularity ◽  
Multi Agent ◽  
Individual Domain ◽  
Behavior Based

Many architectures of mind assume some form of modularity, but what is meant by the term ‘module’? This chapter creates a framework for understanding current modularity research in three subdisciplines of cognitive science: psychology, artificial intelligence (AI), and neuroscience. This framework starts from the distinction between horizontal modules that support all expressed behaviors vs. vertical modules that support individual domain-specific capacities. The framework is used to discuss innateness, automaticity, compositionality, representations, massive modularity, behavior-based and multi-agent AI systems, and correspondence to physiological neurosystems. There is also a brief discussion of the relevance of modularity to conscious experience.

Rule Engines and Agent-Based Systems

2012 ◽  
pp. 211-218 ◽  
Author(s):  
Agostino Poggi ◽  
Michele Tomaiuolo
Keyword(s):  
Dynamic Environments ◽  
Multi Agent Systems ◽  
Rule Engine ◽  
Rule Based ◽  
Agent Based ◽  
Agent Systems ◽  
Loosely Coupled ◽  
Domain Specific ◽  
Multi Agent ◽  
Good Support

Expert systems are successfully applied to a number of domains. Often built on generic rule-based systems, they can also exploit optimized algorithms. On the other side, being based on loosely coupled components and peer to peer infrastructures for asynchronous messaging, multi-agent systems allow code mobility, adaptability, easy of deployment and reconfiguration, thus fitting distributed and dynamic environments. Also, they have good support for domain specific ontologies, an important feature when modelling human experts’ knowledge. The possibility of obtaining the best features of both technologies is concretely demonstrated by the integration of JBoss Rules, a rule engine efficiently implementing the Rete-OO algorithm, into JADE, a FIPA-compliant multi-agent system.

Multi-Agent Architecture for Knowledge-Driven Decision Support

2010 ◽  
pp. 28-46
Author(s):  
Rahul Singh
Keyword(s):  
Decision Support ◽  
Intelligent Agents ◽  
Agent Architecture ◽  
Specific Knowledge ◽  
Domain Specific ◽  
Knowledge Based ◽  
Intelligent Software Agents ◽  
Intelligent Decision ◽  
Domain Specific Knowledge ◽  
Multi Agent

Organizations use knowledge-driven systems to deliver problem-specific knowledge over Internet-based distributed platforms to decision-makers. Increasingly, artificial intelligence (AI) techniques for knowledge representation are being used to deliver knowledge-driven decision support in multiple forms. In this chapter, we present an Architecture for knowledge-based decision support, delivered through a Multi-Agent Architecture. We illustrate how to represent and exchange domain-specific knowledge in XML-format through intelligent agents to create exchange and use knowledge to provide intelligent decision support. We show the integration of knowledge discovery techniques to create knowledge from organizational data; and knowledge repositories (KR) to store, manage and use data by intelligent software agents for effective knowledge-driven decision support. Implementation details of the architecture, its business implications and directions for further research are discussed.

scholarly journals Few-Shot Bayesian Imitation Learning with Logical Program Policies

2020 ◽  
Vol 34 (06) ◽  
pp. 10251-10258
Author(s):  
Tom Silver ◽  
Kelsey R. Allen ◽  
Alex K. Lew ◽  
Leslie Pack Kaelbling ◽  
Josh Tenenbaum
Keyword(s):  
Learning Algorithm ◽  
Training Data ◽  
Policy Learning ◽  
Domain Specific Language ◽  
Computationally Efficient ◽  
Domain Specific ◽  
Strategy Games ◽  
Approximate Bayesian ◽  
Data Requirements ◽  
Approximate Bayesian Inference

Humans can learn many novel tasks from a very small number (1–5) of demonstrations, in stark contrast to the data requirements of nearly tabula rasa deep learning methods. We propose an expressive class of policies, a strong but general prior, and a learning algorithm that, together, can learn interesting policies from very few examples. We represent policies as logical combinations of programs drawn from a domain-specific language (DSL), define a prior over policies with a probabilistic grammar, and derive an approximate Bayesian inference algorithm to learn policies from demonstrations. In experiments, we study six strategy games played on a 2D grid with one shared DSL. After a few demonstrations of each game, the inferred policies generalize to new game instances that differ substantially from the demonstrations. Our policy learning is 20–1,000x more data efficient than convolutional and fully convolutional policy learning and many orders of magnitude more computationally efficient than vanilla program induction. We argue that the proposed method is an apt choice for tasks that have scarce training data and feature significant, structured variation between task instances.

INTENT RECOGNITION IN MULTI-AGENT SYSTEMS: COW HERDING

2014 ◽  
Vol 28 (05) ◽  
pp. 1459006
Author(s):  
NAJLA AHMAD ◽  
ARVIN AGAH
Keyword(s):  
Conceptual Framework ◽  
Recognition System ◽  
Plan Recognition ◽  
Multi Agent System ◽  
Multi Agent Systems ◽  
Agent System ◽  
Intent Recognition ◽  
Domain Specific ◽  
Novel Approach ◽  
Multi Agent

In a multi-agent system, an agent may utilize its idle time to assist other agents in the system. Intent recognition is proposed to accomplish this with minimal communication. An agent performing recognition observes the tasks other agents are performing and, unlike the much studied field of plan recognition, the overall intent of an agent is recognized instead of a specific plan. The observing agent may use capabilities that it has not observed. A conceptual framework is proposed for intent recognition systems. An implementation of the conceptual framework is tested and evaluated. We hypothesize that using intent recognition in a multi-agent system increases utility (where utility is domain specific) and decreases the amount of communication. We test our hypotheses using the domain of Cow Herding, where agents attempt to herd cow agents into team corrals. A set of metrics, including task time and number of communications, is used to compare the performance of plan recognition and intent recognition. In our results, we find that intent recognition agents communicate fewer times than plan recognition agents. In addition, unlike plan recognition, when agents use the novel approach of intent recognition, they select unobserved actions to perform. Intent recognition agents were also able to outperform plan recognition agents by consistently scoring more points in the Cow Herding domain. This research shows that under certain conditions, an intent recognition system is more efficient than a plan recognition system. The advantage of intent recognition over plan recognition becomes more apparent in complex domains.

Off-policy learning for adaptive optimal output synchronization of heterogeneous multi-agent systems

Automatica ◽  
2020 ◽  
Vol 119 ◽  
pp. 109081
Author(s):  
Ci Chen ◽  
Frank L. Lewis ◽  
Kan Xie ◽  
Shengli Xie ◽  
Yilu Liu
Keyword(s):  
Policy Learning ◽  
Multi Agent Systems ◽  
Agent Systems ◽  
Optimal Output ◽  
Output Synchronization ◽  
Multi Agent

scholarly journals Task-oriented Domain-specific Meta-Embedding for Text Classification

2020 ◽  
Author(s):  
Xin Wu ◽  
Yi Cai ◽  
Yang Kai ◽  
Tao Wang ◽  
Qing Li
Keyword(s):  
Text Classification ◽  
Domain Specific ◽  
Task Oriented
Sign in / Sign up

Export Citation Format

Share Document