Over the month or so, I have been introduced to the secrets in the world of Autonomy technology as part of a job for a business partner. My first encounter of Autonomy as a company dates some years back, still I haven’t stumbled upon the name on the web since then even though I tend to crawl all sorts of sites for interesting technology content. Now after being pulled even closer to the actual technology and start learning about the cool products they make, I have been searching the web to learn more about Autonomy and their product portfolio. Especially, I have been looking for technical articles on how to solve certain customer situations or making cool scripts to do certain things. All the stuff that you normally find when dealing with a software product with that much advanced technology and functionality. Well, I might need to take a step back. What is Autonomy?

Autonomy is a humongous company dealing with meaning based computing. It’s a Cambrigde University spin off product and their core technology is their IDOL server which is able to understand of kinds of information. Yes, understand or perhaps more correct, have the ability to know how much things are connected despite them being unstructured without explicit semantic relationship. Its algorithms are based on measuring distance between letters, words, sentences etc. Not distance as in kilometers, but in terms of a virtual distance. The exact algorithms and technology to do this is naturally a well kept business secret. However, I came across different scientific documents while I was writing my master thesis touching upon ubiquitous computing which ties more or less directly into the requirements for measuring these kinds of distances in a multidimensional environment. Note that I do believe that measuring the distance is the only thing needed in order to make a product like IDOL. This way of identifying relevance makes IDOL a great search engine in some scenarios. Normal search engines do keyword searches, which mean that "cat" and "cow" is not associated at all, however, IDOL will recognize that both are animals and therefore will have a bigger relevance that for example "cat" and "book".

Autonomy - An introduction to meaning based computing

Autonomy websites are corporate, very corporate. All of them and that is the issue. It’s almost impossible to get some more in depth information about the configuration with it being turned into marketing and sales speeches. At least that’s the stuff I have been met with. I’m currently working with Autonomy Etalk (also called Qfiniti) and even that is actually a suite of products. So Qfiniti Survey (which has the ability to survey customer calling your call center in order for them to rate the service etc.) is more correct. As this is a European installation we are working with E1 telephony lines and Survey is targeted towards T1 (US standard) we are moving between documented and undocumented setup procedures and even log entries are hard to find documentation for. When you encounter a mystified log entry, you would most of the time search the internet hoping others might have the same issue and has found a solution, but this is not possible with Autonomy stuff. At least the issues I have been dealing with until now.

Even though most Qfiniti installations differ greatly due to the different technical environments, some issues have to be similar or at least reading about similar issues might help resolve them. If appropriate I will in the future disclose some of the issues we have run into and how we solved them.

Automony, Qfiniti, Etalk, IDOL, meaning based computing