[sc34wg3] Are topic maps about knowledge representation?

[Kal Ahmed]

Steve Pepper wrote:

> In the current "Editor's working copy" of the SAM at
>
>    http://www.isotopicmaps.org/sam/sam-model/
>
> the Introduction starts with this paragraph:
>
>   Topic maps are abstract structures which encode
>   information about a domain and connect this information to
>   information resources that are considered relevant to the
>   domain. Topic maps are organized around topics, which are
>   symbols representing subjects of discourse, associations
>   representing relationships between the subjects, and
>   occurrences, which connect the subjects to pertinent
>   information resources.
>
> I propose to change this to the following:
>
>   Topic maps are abstract structures which encode
>   knowledge about a domain and connect this knowledge to
>   information resources that are considered relevant to the
>   domain. Topic maps are organized around topics, which are
>   symbols representing subjects of discourse, associations
>   representing relationships between the subjects, and
>   occurrences, which connect the subjects to pertinent
>   information resources.
>
> To me, topic maps *are* about knowledge representation and I
> think we should say so clearly in the standard.
>
> I strongly believe there is a sense in which we can legitimately
> claim that topic maps capture knowledge. (For the record, I also
> believe that we cannot legitimately say that topic maps capture
> knowledge in ALL senses of the word...)

I agree that there is some sense in which we could claim that topic maps 
are a tool for knowledge representation. But I don't think that it is 
really an accurate claim for a couple of reasons:

1) Knowledge is only in the knower. So an encoding of knowledge becomes 
information - only when it is decoded (by a human) can it become 
knowledge again. This means that the second occurrence of the word 
knowledge should really be "encoded knowledge" or "information".

2) The word knowledge could be too narrow! For example, I can create a 
topic map that reflects the data held in multiple relational databases; 
or a topic map that details the contents of my email in-box. Neither of 
those things are knowledge. This means that the first occurrence of the 
word "knowledge" is too limiting.

IMHO "knowledge" is either a philosophical term or a buzz-word. One 
route leads to 2000 years of epitomological debate and the other leads 
to a misunderstanding and a pigeon-holing of the technology. In either 
case it should be avoided in a standard.

Cheers,

Kal