Knowledge Structuring and Artificial Intelligence

Knowledge Structuring and Artificial Intelligence

Postby kyletheninja » Tue May 06, 2014 9:27 pm

I am extremely happy to see the FreePlane team taking off from the rather static status of our beloved FreeMind. As some one who sees incredible possibilities coming out of the knowledge-representation industry in the near future, I am overjoyed to see these Free Software gurus hard at work. Here I would like to outline not only why I think applications like FreePlane could become involved in new-age artificial intelligence applications, but how specifically this can be achieved.

:ugeek:

- Concept Mapping is what relates to artificial intelligence most directly. The Cmaps team is doing really interesting things with AI already that I think could make use of their concept mapping faculties if they don't already (http://www.ihmc.us/Research/human_machine.php). The reason I think mindmapping is so important, is because it is more practically applicable to finding solutions to real-life problems for normal people.

What I have been interested since I came up with the idea of mindmaps myself (before I found all of the beautiful software that was already out there) is in giving normal people the ability to causally and even computationally model real-life phenomenon in a casual way that helps them make decisions and perform tasks. Below I list the dream features I would have added to FreePlane in order to approach this and similar goals.

1) Messages between nodes. Every node has automatic (hidden when empty) subnodes "inbox" and "outbox". Then by selecting "message" from a node's context menu (or whatever) and then selecting another node(s) a user can make a comment about how any two or more concepts relate.

2) Relations attached to messages. In the message box that pops up a user can select a relation from the relations menu. Those relations are defined by being children of the relations node.

3) (Optional) Display the relationships formed by relation labelled messages between nodes in concept map form (directional graph).

I will save any further explanation of my ideas (pages upon pages friend) for the event that anyone cares, but I want to make it clear. I am a recent graduate who has studied mathematics, psychology, statistics and -from more theoretical than applied perspectives- computer science fairly extensively, and while I cannot say I know for certain or have proof, I truly believe that this is the path towards the type of AI that can interact with human beings on a natural level. That is, it could know everything we can map, and answer our questions according to all of it.

I think the best open source mindmapping software out there will have a role to play in this coming about in a positive way.
kyletheninja
 
Posts: 2
Joined: Tue May 06, 2014 8:03 pm

Re: Knowledge Structuring and Artificial Intelligence

Postby dpolivaev » Wed May 07, 2014 2:16 pm

Hi,

I am interested in making possible new ways to visualize data relationships in mind maps. And I also like helping people to realize their ideas. So I would like to suggest two ways how you could add your message concept to Freeplane.

1. For better understanding of your idea I want to be able to see some concrete use cases behind it: what kinds of data do need this inbox-outbox approach?
2. When I think I know what you mean I could think about the solution in code. Would you like to contribute to the implementation yourself? I think once you get into the project you could experiment with other ideas too.

Regards, Dimitry
User avatar
dpolivaev
Site Admin
 
Posts: 667
Joined: Sat Jul 23, 2011 1:21 pm

Re: Knowledge Structuring and Artificial Intelligence

Postby endlessloop » Thu May 08, 2014 7:43 pm

kyletheninja wrote: - Concept Mapping is what relates to artificial intelligence most directly. The Cmaps team is doing really interesting things with AI already that I think could make use of their concept mapping faculties if they don't already (http://www.ihmc.us/Research/human_machine.php). The reason I think mindmapping is so important, is because it is more practically applicable to finding solutions to real-life problems for normal people.


Freeplane is a tool dedicated to mind-mapping not concept-mapping. There was a discussion on this forum (in the thread "NEW FEATURES LIST +++ PLEASE") about enhancements which will be able to transform Freeplane to a (good) tool on concept-mapping by adding the possibility to have several roots, free-nodes more easy, etc. But the answer was : "no". And the reason was that Freeplane is for mindmapping only and it is not stupid, IMHO. Because there is a very good tool to make concept-maps : Visual Undestanding Environment ("VUE", see your search engine, it is developped by TUFT university).


kyletheninja wrote:1) Messages between nodes. Every node has automatic (hidden when empty) subnodes "inbox" and "outbox". Then by selecting "message" from a node's context menu (or whatever) and then selecting another node(s) a user can make a comment about how any two or more concepts relate.

2) Relations attached to messages. In the message box that pops up a user can select a relation from the relations menu. Those relations are defined by being children of the relations node.

3) (Optional) Display the relationships formed by relation labelled messages between nodes in concept map form (directional graph).


You have that (and much more) in VUE. VUE is not a good tool for mind-mapping (I think Freeplane is the best for that) but it is the best tool for what you want (concept-mapping and visual thinking).

Regards.
endlessloop
 
Posts: 12
Joined: Sun Jan 05, 2014 8:19 pm


Return to Open Discussion

Who is online

Users browsing this forum: No registered users and 1 guest

cron