« Hello? Topic Maps, Hello? | Main | Survey on Search, Metadata, and Taxonomy Practices »

August 24, 2005

The Trouble with Topic Maps

We have got to get better and communicating technology to the masses. I posted a comment about topic maps yesterday where I got some great responses with links. These links and most of the other references on topic maps work very hard to make something as simple as a book index into something complicated. HTML succeeded because it was simple, easy to use, and tools emerged where a five year old could build a web page. So let us review; HTML, simple and easy equals mass adoption while topic maps, complicated and confusing equals death. Topic maps are a great idea and I can only imagine what it would be like to have an actual index of the web available but we had better start simplifying things so that the majority of people can understand the value and utility. Still disagree? Take a look at Google interface and try to describe the recursive search algorithm in a single sentence. The interface is simple while the complexity is taken out back to the shed…

Posted by Todd at August 24, 2005 11:56 AM

Trackback Pings

TrackBack URL for this entry:
http://www.rtodd.com/cgi-bin/mt/mt-tb.cgi/10

Comments

Html simple? And what about xhtml? or xhtml 2?

Maybe (x)html isn't so simple, but there are a lot of instrument to work with :)

Posted by: Salvatore at September 4, 2005 11:40 PM

Html simple? And what about xhtml? or xhtml 2?

Maybe (x)html isn't so simple, but there are a lot of instrument to work with :)

Posted by: Salvatore at September 4, 2005 11:41 PM

I think there are two main problems with your arguments here:

1. You are comparing different levels of abstraction
2. You mix ‘end user’ with developer

Designing a Topic Map is comparable to designing a database or ontology. Comparing it to making an html web page or using google is missing the point completely.

As a matter of fact Topic Maps is one of the metamodels included in the OMG 'Ontology Definition Metamodel' RFP:
http://www.omg.org/ontology/ontology_info.htm

> We have got to get better and communicating technology to the masses.

Considering that you use "a five year old" and "the majority of people" in your examples, it seems like your are talking about end users, not developers.

Why should an end user care what kind of base technology the application use? You have to differentiate between developers and end users.

The developer has to choose the appropriate technology to use, and the end user may judge the result. A book index, for instance, is simple to use, but can be hard to construct. The user doesn’t have to know anything about how to make an index however.

Your average Google user wouldn't want to write html tags, which was necessary the first years of building the web. In the case of HTML, the technology was easy for any developer to code and error tolerant.

Most of the Topic Maps based web applications I know of look like any well designed web application to the user. Just like the Google interface looks like any search engine. - Which is probably why you haven't noticed the use of the TM technology lately.

Most Topic Maps based applications have a standard forms-based web interface for end user editing, just like any normal web application.

>tools emerged where a five year old could build a web page

It’s tempting to argue here, but let’s say we agree that most five year olds don’t read and write, and use the term “school kid”.

The Norwegian company Cerpus have made the application Brainbank, which is made to be used by school kids:
http://cerpus.com/

Posted by: Are Gulbrandsen at September 5, 2005 10:10 AM

Are made very good points; end-user != developer.

If complexity is the problem; why aren't object- / relational database systems, or object oriented programming (where the programmer also constructs (some sort of) an ontology) dead. And compared to topic maps; why isn't W3C's OWL gone?

Everything is not supposed to be designed for, or used by, "the masses". As a side note: if you read Berners-Lees visions for the Semantic Web, it's not about porting every (X)HTML page out there to OWL/RDF(S) (or topic maps for that sake - although their primary use is not SW, it sure is a good way to go).

Posted by: teepee at September 5, 2005 12:54 PM

Copyright © 2002 - 2005 - R. Todd Stephens, Ph.D. All rights reserved.