unconstrained retrievals

[Christian Lebiere]

> Instead, I use unconstrained retrieval that simply returns the most
> active chunk:
>
> (p retrieve-pattern
>   =goal>
>   ...
> ==>
>   +retrieval>
>     isa pattern
>   =goal>
>   ...
> )
>
> Now the problem with this kind of retrieval is that it often returns
> chunks that are already elements of the goal (because of the strong
> self-activation of chunks), which is of course undesired in the
> present context. Does anybody know a good solution to that problem or
> is there even a "standard" solution?

The associative link from a chunk to itself might be strong if it it 
appears in few or no other chunks (default equation) or if it is often 
directly retrieved (learning equation), but it is not in any way 
priviledged over other links.  So this is really an instance of a more 
general problem of relying only on activation for retrieval.  Unconstrained 
retrieval is essentially free association.  It happens sometimes, but it is 
not the only or even probably the most common form of memory access.

Christian