Re: LIDA Framework

[Ryan J. McCall]

Hi Marcin,

Nice to meet you. It's good to hear about your interest in LIDA. I would suggest, if you haven't already, to read the LIDA tutorial introducing the framework:

http://ccrg.cs.memphis.edu/assets/framework/The-LIDA-Tutorial.pdf

My second suggestion is to work through the LIDA framework tutorial that was given at the Artificial
General Intelligence 2011 conference (AGI-11). This should help familiarize you with using the framework. Note that this tutorial is
designed specifically for the older LIDA framework version 1.1b and is not 100% compatible with other framework versions.

You can download the AGI-11 tutorial here:

http://ccrg.cs.memphis.edu/assets/code/lida-framework-tutorial-v1.1b.zip

Best,

Ryan


On 1/28/14, 4:11 AM, Marcin Hernes wrote:

Dear LIDA experts,

My name is Marcin Hernes. I am research scientist and Java programmer at the Wrocław University of Economics, Poland (http://www.ue.wroc.pl/en/). I am interested in using the LIDA framework in the field of economic informatics (Decision Support Systems, Integrated Management Information Systems etc.). First I would like to use the LIDA agent to process users’ opinions (they are placed, for example at internet shops, forums, social networks) about the products (e.g. a polarity of the opinion, a good and bed features of the product) . The opinions are written in natural language (polish). I have been reading the LIDA model and framework papers, I have  installed  the framework and run the ALifeAgent 1.2 example.  I am trying to create new agent (on the basis of ALifeAgent). I changed the context of the configuration files, the code of environment package files and featuredetectors package files. My agent is running, but for now it does nothing.  I have a problem, because I don’t know, what are the steps I need to take to create the agent, which can process users’  opinion (for now I do it by trial and error method). I would like to kindly ask you, could you give me any guidelines in this topic?

 

Best regards

Marcin Hernes

---

Ta wiadomość e-mail jest wolna od wirusów i złośliwego oprogramowania, ponieważ ochrona avast! Antivirus jest aktywna.

-- 
Ryan J. McCall
Ph.D. Candidate, Dept. of Computer Science
Cognitive Computing Research Group
Institute for Intelligent Systems
The University of Memphis
ryanjmccall.com

[hernes...@gmail.com]

Hi Ryan,

Nice to meet you, too. Thank you very much for your suggestions. I has been done the excercises of LIDA framework tutorial that was given at the Artificial
General Intelligence 2011 conference. It helped me to undestand many issues of lida framework.  I have two questions:

1. How is the best way to run two or more agents (with different agents configuration files) without creating new java project (I use Eclipse)?
2. What does LATERAL (link category) mean?

 

Thanks in advance,

Marcin

[Ryan J. McCall]

Hi Marcin,

I'm glad to hear that the tutorial is of help.

1. The short answer is that the LIDA framework does not support multiple concurrent agents. One solution is to use the framework as is to implement a main agent, then implement the second (simpler) agent as part of the Environment class. That is to say, the process (task) that updates the Environment would also update the inputs/outputs of the simple agent. Then, the complex LIDA framework can sense and response to the simple agent via the Environment. A second unexplored solution is to actually hack together two framework agents. One would have to adapt the code in the method "private static void run" in AgentStarter. One major issue is that each Agent created by this code contains its own TaskManager object, so you'd somehow have to drive both TaskManagers from one place... There hasn't been much more thought on this issue so be warned there may be other issues!

2. LinkCategory represents the semantic meaning of a link. These primitive LinkCategory objects defined in PerceptualAssociativeMemoryImpl are just default example ones. The framework doesn't do anything special with them. I think the theoretical idea behind LATERAL Was to represent an association based on similarity like red->orange or dog->cat that involves nodes at a similar level of abstraction and thus the connection is lateral in terms of a hierarchy.

Best,

Ryan

--
You received this message because you are subscribed to the Google Groups "Cognitive Computing Research Group - CCRG" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ccrg-memphis...@googlegroups.com.
To post to this group, send email to ccrg-m...@googlegroups.com.
Visit this group at http://groups.google.com/group/ccrg-memphis.
For more options, visit https://groups.google.com/groups/opt_out.