Cambridge Cybernetic Society
9th Meeting
"Searching, Searching:
Knowledge Acquisition by any other Name"
A demonstration of innovative prototype software that asks the question:
Isn't seaching on the Internet just like acquiring knowledge? or
Making browser interfaces to Search Engine obsolete
Grasp Information Corporation
Wednesday 28 May 1997
Grasp Information Corporation
185 Alewife Brook Parkway
Cambridge MA 02138
617-441-9900 x121
5:00 to 7:00 pm
{There is ample parking behind the building. 8 minute walk from Alewife Red Line "T" Station.}
Abstract
Many of us are plagued by unfortunate experiences with searching for information on the Internet (or, coming soon, the intranet). We already know that words and word clusters do not constitute the semantic content of a document, but we also know that, with the current generation of typical search engines, it's the best we can do with the search back-end right now.
But what about the front-end, the user interface to search engines? Surely our search tools can be improved, but how? In a prototype under development at Grasp Information Corporation, the premise is that any search process is fundamentally a learning process and therefore insights about knoweldge acquisition are valuable in designing new user interfaces to the search process. Moving from an undifferentiated field, to user-meaningful distinctions, to re-structuring the domain based on those distinctions are all aspects of learning something new. What are the necessary software functions to handle these two broad categories: understanding an overall domain vs. finding a specific datum?
With an innovative prototype under development, Grasp Information Corporation hopes to revolutionize every-day use of serach engines by providing pratical functions based on the above insights. These functions are not semantically-oriented, or based in complex or unreliable extracting of summarizations from texts. A prototype will be shown that
- eliminates the need to use the typical search engine interface from browser companies
- provides automatic agents that determine the validity of links and reports back with clear, visual cues
- automatically extracts "relevant" snippets and categorizes them
- automatically generates useful distinctions and re-categorizes search results
Attendees are invited to suggest (in advance or during the
meeting) their own search problems.
Meeting Host Paul Pangaro is working with Grasp
Information Corporation, a venture-backed software company
that specializes in creating innovative tools that improve productivity
before, during and after the search.
© Copyright Paul Pangaro 1994 - 2000. All Rights Reserved.