Tuesday, September 21, 2010

Usage of Wolfram|Alpha

Wolfram|Alpha started round about 1,5 years ago with big noise in the IT related press. Their goal to make "systematic knowledge immediately computable and accessible to everyone" (see http://www.wolframalpha.com/about.html) was promising from the beginning.

Up to now they make huge progress and their engine is a sufficient help for everyone using a computer on a daily basis.

Each day you have to answer question requires specific knowledge or algorithm. In many cases you can find specific Websites helping to answer those questions. But having one source, answer those questions, combine the results for as many domains as Wolfram|Alpha do is unique on the Net (AFAIK).

Samples

Once you want to use or share a certain query you can create a so called widget and bookmark it, embed it into your website, .....

creating a wizard to convert any given unix timestamp to a given named timezone (specific query shown above) i created the following widget (it takes me 10 min. of my limited time ;-)



Limitation

There are still lots of limitations, e.g. for many knowledge domains only US based sources are available. The combination of knowledge is still limited, a fancy query, like "weather the day Jimmy Hendrix died" does not work for many scenarios right now. But they getting better each day....

The end of power point....

My previous post (see http://trent-intovalue.blogspot.com/2010/09/end-of-documents.html) shows that a linear sequence of information is not a perfect match for the expression of complex technical content.
That is also true for presentations which are often supported by power point slides. If you already tried to introduce a complex topic and interact with the audience you might now that the right information is always a slide away.

Why all office application re-create power point instead of thinking of more flexible concepts? I'm not really sure. Few days ago i stumbled upon  Prezi which tries to goes a different route. The content is organized in a tree of topics and navigation is much more focused on context / audience and concrete situation.

Sample:

The end of documents....

Assumption

If you have to describe any technical subject you are aware that knowledge is hard to express in a straight linear sequence of information topics. In most cases you have to structure the information into information tropics and semantic connection between those => a network of information topics.

Nothing new and a trivial statement, you might say. All semantic concepts are using those principle buildup.

Yes, but why most technical subjects are still using documents or slide shows to express technical subjects?
Those media formats are linear by design. The reader has to follow the one and only linear flow defined by the author of the document. In the best case the author is able to find one of the sufficient linear paths through the network of information and the reader is therefore able to understand the described subject. But even in this case getting the hole picture, identify ways to extend the provided information, embed it to different subject etc. isn't possible or at least requires to re-construct the information tree in mind.

Ask yourself why you are using
  • A word processing software to define project information (requirements, design specification, test specification, ....)
  • Power Point to introduce a particular problem domain
  • A word processing software to trace a result of a workshop (also known as workshop protocol)
  • ....
The answer is simple. Because we simple get used to and there are no mainstream alternative media formats out there which can be used without at least one significant constrain (effort to implement and train, difficult to share, ....). The complete office suites still remains rooted in the old linear concepts. Even new players in this business adapting this paradigm (e.g. Google Docs).

Alternatives?

I'm pretty sure that in the future documents will be replaced with applications which providing a way to describe topics as short topics and makes it easy to connect those topics with semantic links (e.g. depends on, contains, .....). A document in this scenario is just one path through the network of information for one particular use case.  This kind of application can replace todays word processing software without loosing any important feature.

In the domain of technical writing topic based authoring (e.g. using DITA information architecture) becomes more and more popular. The main use case there is to (re-)use information as much as possible to reduce creation and information maintenance costs. In my point of view that is "only" a important side effect of having the content defined in a much more usable form. Not a linear sequence of information reflecting a group of authors view but as more or less complete set of information topics linked together. The todays results are still linear documents of some format (pdf, online help formats, ....) but that only corroborate the belief that linear documents are mainstream.

Open Issue


The usability of topic based authoring isn't sufficient today. It is more or less a hand crafted creation of enriched information. To dive into the mainstream usability is the most important factor. The creation and linking of information must be at least as easy as using e.g. Mind Mapping tools (e.g. FreeMind, MindManager) combined with easy to use structured topic content editor (e.g. tools like Xopus or XMAX goes into this direction).



There’s more to come? Lets see....