Strategic Foresight

Unbound Spiral: Augmenting Search and Scenarios

August 22, 2003 10:56 AM

A belated post from WFS.  John Petersen is a highly respected futurist. He runs The Arlington Institute and I'm sure many of you know about the value of his bi-weekly newsletter.

John presented LISA: A breakthrough Technology for Anticipating the Future. The LISA tool provide an "information and knowledge engine" that can globally track emerging ideas around any selected topic from thousands of sources. For fellow bloggers... This is about augmenting search capabilities.  It is worth a broader dialogue and you may have to click the links below to get a feel for what's being done. Plus, if you think the collective noosphere and smarts of bloggers will beat the government then we better learn from this!

John is in the business of trying to think about big global futures. He too uses Scenarios. This however is more than a scenario it is being put into practice with millions (i'd guess) spent already. 

This tool is being created for the Singapore Government. They are hoping it will prevent being caught out by SARS like surprises in the future. That's probably just the public side of what's envisaged. It's a systematic tool. (I'm hoping I can capture it here in words.) John is reinforcing that it is being put together with off the shelf tested technologies and that is what's interesting about it.  It's an excellent synthesis or prevailing technologies. There are a few Google listings around it.  The components are listed below. 

Part of the challenge is to design a process that can find, picture, realize, foresee, model and decide (using humans - who do a great job of dealing with complexity) in easy to use interfaces. These components make up LISA's augmented search capability. 

  1. First we are looking at search that uses Convera. So match Google with Convera and you can start searching in multi-languages.  "Convera's unstructured information search across more than 200 forms of text, video, image and audio information, in more than 45 languages. Ie get a planetary view.
  2. Semantix: provides cross-lingual search capability in an off-the-shelf solution that allows users to enter a search request in one language and receive relevant result documents in several different languages
  3. Kapow Technologies has developed an intelligent platform that extracts information from HTML data sources, and converts this into more structured data forms, such as XML and relational databases,
  4. Streamlogic Inc: "Instead of archiving data and running search queries through it, we archive search queries and run data through it," Streamlogic Inc. "It's a search engine on its head." "The advantage of an inverted search engine, is that it's 6,000 times more efficient than the conventional approach. It can handle huge volumes of data that would be expensive or impossible to process using the standard method of loading data into an archive, indexing it and then retroactively querying it."
  5. DSO:  ??? Not sure what this links to. Can't remember. 

You might wonder who would want this apart from scenarist...  Even multinationals will want it.  It's the next level in power searches.  No one will launch a product anywhere without it being recorded or tested in a global context. So once one has the seach function in question up and running how do you use it?  What interface is required? What kind of interfaces will humans use? 

The examples show the latest in displays are being used. 

Starting with Pacific Northwest National Laboratory: Starlight: is a forerunner of an emerging new class of information system, one that couples advanced information modeling and management functionality with a visualization-oriented user interface. This approach makes relationships that exist among the items in the system visible, enabling exciting and powerful new forms of information access, exploitation, and control. Further, Starlight visualization tools employ a common XML-based information model capable of effectively capturing multiple types of relationships that may exist among information of disparate kinds.

Saffron? Add in for additiona concept mapping --- hard to get everything! and apply - Think Tools for developing scenarios and scenario strategies. So imagine a solution that visualizes and leverages vital quantitative input and merges it with qualitative data. Thus ideas, experience, interpretations - even 'gut feel' - become measurable and indispensable - and transparent. Part of the overall analysis.  

So a little repetitively, some of the claims I was hearing this bundling could make... (paraphasing some again...)

Looking at the visualization technology "Starlight" (very cool!) provides visualization of temporal mapping. This provides the capability to look acrosss planetary idea space... then with comparisons you can look at the evolution of ideas... do they become clusters of their own ... how do they migrate.... 

Saffron is a powerful adaptive learning agent based on a dynamic agent so they can predict complex relationships in real time. Saffron observes users online selctions and uses that information to guild future searches. Once it starts to learn basic relationships it gets smarter and smarter... XML documents to Saffron's xml processor transforms information into entities which are then put into the network. Each one with 100k links that can be tied to the network. So then you have a network of the network of all these relationships. Apparently they can do this in compressed form in a real-time manner once you have figured out broadly the kinds of things you are interested in.

For example you can then look for SARs in the last 24 hours.... By using Kapow to provide armies of robots to do the searching the robots will go as often as you want and thus there is no need to retain all the data.  Having set it up you can do it in multiple languages.  

Then extract out of the system where the key nodes are..... with a cluster you can see which are most influencial. Those are the ones that are the most interesting.... This can shift using scenarios. SCAN data, FIND patterns, PICTURE mental models, VISUALIZE, Generalize, Guide behavior adjust by bulding a three screen cockpit.  Anticipate surprises using this technology.

1 Build surprise scenarios. If you don't think about these things before hand it doesn't do any good. 

2. Identify prerequisite events. There have always been recursor indicators we just don't know we should be looking or where.

3. Monitor the recursor indicators proposing to make scenarios a real discipline. Take it down one level... build a scenario database... events bounce off the past; bounce it against the future see the scenarios it is related to then bouce it against the present ..... So surround the event ... with everything you know.

For me... this made being at the WFS session worthwhile.