Canonical reference on agent-based computing

Question

I am interested in exploring the world of BDI agents (software agents that possess "beliefs, desires, intentions", essentially the agent has knowledge of the world, a set of motivations, and carries out certain plans).

I recently read A Canonical Agent Model for Healthcare Applications [1], which left me with a lot of questions, particularly about the specialization of different agent models for particular applications.

The particular modeling language used in their examples was ProForma, and I understand that this is more for the abstract specification of an agent, and that something like 3APL can be used as an actual programming language in this regard, with syntax like:

BELIEFBASE {
    status(standby).
    at(0,0).
    location(r1,2,4).
    location(r5,6,1).
    dirty(r1).
    dirty(r5).
}

My question is, all of these systems clearly reflect years of cumulative efforts, and rather than jumping in to the deep end, I'd like to ease into this world of research a bit more slowly. Is there a canonical reference in this area that might be able to provide a more general overview of all of these levels of organization, and where the abstractions stop and the implementations begin?

---

1. Fox J., Glasspool, D., Modgil, S. A Canonical Agent Model for Healthcare Applications. IEEE Intelligent Systems, 21(6), 21-28, 2006.

Answer 1

If you want to approach this field from a computer science perspective then the standard reference I would recommend is:

Yoav Shoham and Kevin Leyton-Brown [2009], "Multiagent systems: algorithmic, game-theoretic, and logical foundations", Cambridge University press.

Currently, it seems like the focus of theoretical work in this field (and this is obviously biased by my own interests) is to consider environments where agents are limited in their interactions. If you just have agents meet randomly, then ABMs are usually overkill and things can be done analytically. However, if there is some interesting network structure (fixed or dynamic) to the interactions (as there often is in real life) then ABMs become an essential tool. A good book discussing some of these ideas from a CS perspective is:

David Easley and Jon Kleinberg [2010], "Networks, crowds, and markets: Reasoning about a highly connected world," Cambridge University press. (draft available online)

Questions of interest:

• Simulation modeling of diseases

• Sources for Algorithmic Evolutionary Game Theory