How humans want to find their answers

[Robert Hillard]

Having taken the time to get a handle on your enterprise information by roughly dividing it into the four layers I described in my last blog post, I’d like to give you an idea for a “quick win”.

Humans don’t differentiate between the things that are structured (such as the content of the corporate data warehouse) and those that are unstructured (such as PDF train timetables).  While technologists think it is astoundingly obvious that the former is instantly analysable and the latter is only to be read, the same assumption is not obvious to most end-users of the data.

Consider for a moment the new class of analytical search engines such as WolframAlpha and compare to a more traditional tool such as Google.  Try the query “American Airlines on-time performance” in both and compare the results.  WolframAlpha will actually try to give you the answer whereas Google will tell you where to go and find the answer.

If you invest in understanding your content, you can do the same for your organisation, providing tools to your users which will give them the answers to their questions rather than simply telling them “the answer is out there”.

For further reading on this topic, I recommend MIKE2.0’s enterprise search offering




More...

[Doug Heywood]

Interesting blog entry Robert - thanks.

I have tried WolframAlpha and I found that it was only capable of answering a very limited class of analytic questions and that when it doesn't understand the question, it is pretty useless.

So I guess that at the moment, wouldn't an engine/application combination of the two be most effective? What I mean is a search engine that:

1. First attempts to answer the query using an Analytic Engine, and if that doesn't work;
2. Secondly uses a 'google' type search.

So wouldn't I get the best results using something like dogpile once they include Analytic Engines?

Anyways, thanks for the blog - keep it coming!

[Robert Hillard]

Doug,

I agree with your comments. I also think the limitations of these very early analytical search engines show the importance of good metadata and the fact that information engineering after the event is never as effective as properly designed data repositories.

The key point is that people seem to like to consume information using verbs and nouns rather than through abstract analysis.

Cheers,
Rob

I am a firm believer that one can find all the info they need to know about a company or a style of business by reading blogs. Navigating the information management maze could begin by understanding the corporation. If you are new to blogs, especially corporate blogs, then a good starting spot could be blogging for dummies. Understanding blogs and how to get the most out of them is the first step in understanding how both corporations and their information management systems work.

hiii

While technologists think it is astoundingly obvious that the former is instantly analysable and the latter is only to be read, the same assumption is not obvious to most end-users of the data.

_____________________
Earn an Extra $1000 to $1200 per month doing Part Time Data Entry Jobs! Work from home data entry jobs to post simple data submissions on Internet. Make $1 per entry. Easy form filling, data entry and ad posting jobs. No selling, No phone calls, No Marketing. No Investment. Bi-weekly payments. Full Training Provided. Pls visit: Data-Entry

hiii
While technologists think it is astoundingly obvious that the former is instantly analysable and the latter is only to be read, the same assumption is not obvious to most end-users of the data.


________________
Earn an Extra $1000 to $1200 per month doing Part Time Data Entry Jobs! Work from home data entry jobs to post simple data submissions on Internet. Make $1 per entry. Easy form filling, data entry and ad posting jobs. No selling, No phone calls, No Marketing. No Investment. Bi-weekly payments. Full Training Provided. Pls visit: Data-Entry