November 20, 2006
Ilya Geller took part in NIST TREC Conference 2006. He presented his theoretical ontological justification of Differential
Linguistics. Now we have both practical and theoretical proves of our technology absolute superiority!
August 18, 2006
We continue to develop the method that is based upon obtaining of redundant information about the Egoism of
the person seeking the information, and equally redundant information about the Egoism of the creators of
the texts in which the information is being sought.
Ilya Geller is trying to publish
that describe our technique.
January 2, 2006
We participated in NIST TREC QA 2005.
Ilya Geller invites those who want to discuss his ideas to come at
in both Russian and English.
June 1, 2005
Yahoo Offers New Y!Q Contextual Search Tool
Contextual Search with Firefox
Feburary 21, 2005
We started New Technological Line. Unable to get financed we began to imply Principles of Differential
Linguistics and New Mechanics to Information Retrieval
October 17, 2004
We begin the mapping of the Internet.
October 15, 2004
We found that Google is ready to lunch personal Internet Vehicles - http://searchenginewatch.com/searchday/article.php/3421651.
October 1, 2004
We start "Legal Industry" Progect
September 30, 2004
We start "Gene" Progect
July 11, 2004
We finished to test a first personal vehicle for the Internet.
June 26, 2004
We are finishing to test a first personal vehicle for the Internet.
May 2, 2004
We got a response from
National Science Foundation
on our proposal. Its a first officially valid comeback!
February 24, 2004
We found that our technology is stolen at the first time.
November 21, 2003
We took 3-rd place in QA NIST TREC and 9-th in Novelty NIST TREC. Therefore, a search part of our technology works
even if it's far from completion
March 1, 2003
A new technology for AI is tested: we transform the found information in a new quality.
January 14, 2003
We have activated a prototype of STARC, an Artificial Intelligence being, which can support a
primitive but meaningful conversation.
December 17, 2002
Our research has paid off. Clones developed by our software engineers have "replied" with
July 18, 2002
LexiClone succeeds in developing and testing the first Clone application, which imitates conversations
with well known celebrities and historical figures.
July 17, 2002
We begin to test Clones.
June 30, 2002
New Case Studies - for CNN and Delphion - are placed on the site.
June 26, 2002
LexiClone relaunches with new site and an impressive line of new software products. Visit
April 30, 2002
A new release of LexiData, version 1, has been released in beta.
April 29, 2002
Two Case Studies is ready Case Study Patents and Case Study Digesting. The first
compares output list from U.S. Patent and Trademark Office Patent Full-Text and Full-Page
Image Databases and LexiClone filtering of these results; the second creates digests
of any entered document. The results of output list from U.S. Patent and Trademark Office
filtering confirm that there are no patents that work strictly with the Internet language
oriented search engines. Or, that more is the truth, four separated words as the search
string do not work properly.
April 23, 2002
Clones of celebrities were tested using the local program.
March 11, 2002
New hires have been announced. The company has been relaunched as LexiClone.
March 08, 2002
We finished the testing of UniSearch-3.1
January 11, 2002
A new release of UniSearch, version 3, has been released in beta. The search runs faster: a
database of the size identical to previous tests comes back 67 times faster.
December 23, 2001
The concept of human cloning based on lexical particulars has been hatched. Test clones are
created. An average clone answers a search question in 15-20 minutes.
December 12, 2001
Digesting technology has been optimized for the second time, and works faster. A 100 kb document
file has been processed in 20 seconds, using the same computer equipment (Pentium 4, 1.1G,
256 Mb RAM platform.)
November 23, 2001
Corporate Web site has been redesigned. A beta version of a new application, UniSearch-2.1, is
now available for downloads. A small database of documents has been created.
September 03, 2001
The development of UniSearch digesting technology is completed. Digests summarize main
ideas of the corresponding documents. The largest document files to date, ranging from 1.9 to 2.4 Mb,
have been created. An average time to create a digest is between 12 minutes and 18 minutes
(on a Pentium 4, 1.1G, 256 Mb RAM platform).
July 07, 2001
The pilot version of UniSearch-1.9 has been completed. The search now runs on a database of
10,000 separate documents. The search time for a database of 10,000 documents with an average size of
50 kilobits is 6.4 minutes.