The Artificial Knowledge Interface for Reasoning Applications (AKIRA) project aims to create a C++ development framework to build cognitive architectures and complex artificial intelligent agents featuring KQML, fuzzy logic, neural networks, fuzzy cognitive maps, and DIPRA. DIPRA is a distributed version of the BDI (Belief Desire Intention) goal oriented model.
PED is a dialogue management system that uses a probabilistic nested belief model to choose dialogue strategies. The dialogue system designer need only supply a set of plan rules to PED as a dialogue grammar with preconditions. Using this grammar, PED constructs game trees (like the one below) to represent the outcomes of the dialogue, so that a dialogue strategy can be chosen for the current turn in the dialogue. PED automatically maintains a belief model by a belief revision process that uses the observed acts in the dialogue. The game tree is evaluated in the context of this belief model. PED is efficient because it uses probabilistic estimates of belief rather than a plain logical belief model.
Amiba is a Gene Expression Programming (GEP) framework for Java. GEP is, like genetic algorithms, a branch of evolutionary computing. The framework separates the process of evolution from the process of interpretation of the chromosome, allowing the use of various schemes. For example, graphs may be used as terminals and graph operations as operators in the chromosome instead of the usual double precision numbers. It implements mutation, transposition, and recombination. Options and rates are easily configured through an XML file. A mechanism to load fitness cases in bulk is also provided.
JASA is a high-performance auction simulator. It is designed for performing experiments in agent-based computational economics. It implements variants of the double-auction market, which is commonly used to run real-world market places such as stock exchanges. It is designed to be highly extensible so that other types of auctions can easily be implemented. The software also provides a base classes for implementing simple adaptive trading agents.
tinyMAS is a multiagent platform that provides base concepts (such as kernel, message, yellow pages, white pages, and transport service) and extended concepts (such as environment, influence, and perception). It aims to provide an easy-to-understand and and easy-to-use platform dedicated for multiagent engineer/research courses. TinyMAS is no longer under development. A large amount of its source code has been merged into the Janus platform.