That Condor moment

Written By:
Published:
Content Copyright © 2005 Bloor. All Rights Reserved.

There used to be a television advert for a pipe tobacco called Condor. During the course of the advertisement the actor would fill his pipe, lean back in his chair, light the pipe and then, in a tone of complete contentment, the voiceover would say, ah, that Condor moment.

In this industry I don’t often get Condor moments but I had one recently: for the first time, in more than thirteen years as an analyst, I actually looked at a product that I, personally, would buy. I should be clear about this: I have to have an operating system; I have to have word processing; and I have to have various other things – but this is the first product, which I don’t have to have, that I would buy. Note that this doesn’t mean that DB2, say, isn’t any good, it just means that DB2 would not help me to do my job, whereas Summarise! will.

Summarise! is one of several products that are developed and marketed by Corpora Software, which is a dual nationality organisation that comprises both UK- and US-based companies, though Corpora was founded in the UK initially. Unusually for a start-up, the company listed on AIM (alternative investment market) early in its life, primarily so that it was in a position to make acquisitions – of which it has already made several – the most notable of which was Exago, which was spun out of BT and, as result, BT is a major shareholder in Corpora.

The Corpora product set includes:

  • Find!, which is the company’s natural language search engine (though it also uses statistical techniques – it is primarily Bayesian but also has Boolean capabilities). Currently this is only in English though the company is planning an Italian version and thereafter one in French. One notable feature is that security is integrated at the operating system level so that you don’t need any application level logic: in other words, when you search, Find! pulls back all the relevant items that you are permitted to see.
  • Summarize!, which does what it says on the tin. I like it because it plugs into Outlook (and other Microsoft products such as SharePoint – other products can similarly integrate) and you can use it to summarise, and pull out keywords, from any incoming documents (Word, pdf, html and others, even PowerPoint slides) that may be attached to an e-mail. You can state how long you want the summary to be and you can also apply filters for relevancy. You can also summarise against a URL.
  • Jump! will automatically index (and sub-index by context) any document or you can index across multiple documents. One of the neat features (which the company believes to be unique) is that it incorporates pronoun resolution. So you might have an article about George Bush, say, in which he is mentioned by name just once but referenced 13 other times as he and the software will recognise that this is actually 14 references to the US president. Again, you can filter. While the Corpora products do not require you to define a taxonomy, you can use Jump! to help to create a taxonomy if you want to do this, because it extracts subject-verb-object triplets.
  • Burst! is an event aggregation engine. That is, it will access appropriate news feeds, say, and pull out all references to a specified topic over a particular period of time.
  • Sentiment! works in conjunction with Burst! and provides automated sentiment analysis. That is, it will look through the extracted news items (or whatever) and see if the attitude expressed was positive or negative, which will be useful in a variety of situations: for example, if considering an acquisition target.
  • Finally, Connect! Is a portal product for communities of interest and knowledge sharing. This product was originally developed by BT Exact and then acquired by Corpora. It has Find! built into it.

I think that’s a pretty interesting product set. Speaking purely personally, Summarize! will save me time with all the incoming e-mails that I get, it should speed up my research efforts when I am looking at various web sites, and it will help me to check the various documents that I write for their relevancy. Neat stuff.