<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
<style type="text/css" style="display:none;"><!-- P {margin-top:0;margin-bottom:0;} --></style>
</head>
<body dir="ltr">
<div id="divtagdefaultwrapper" style="font-size:12pt;color:#000000;font-family:Calibri,Helvetica,sans-serif;" dir="ltr">
<div><br>
</div>
<div>As Ryan Naraine has pointed out I never did an announcement on this mailing list when
<a href="https://www.cyxtera.com" class="OWAAutoLink" id="LPlnk399062" previewremoved="true">
Cyxtera</a> and Immunity finally closed our deal. Partially that's because these things are in some ways anti-climactic, and partially because I and a lot of the team at Immunity immediately went on a binge of experimenting with various large toolkits we'd
never had access to before.</div>
<div><br>
</div>
<div>For example, this one:</div>
<div><br>
</div>
<div><br>
</div>
<img size="521733" contenttype="image/png" id="794840_rs|3" style="user-select: none; width: 557px; height: 359.138px;" tabindex="0" width="557" height="359" data-outlook-trace="F:1|T:1" src="cid:b6710e97-06db-48bc-a941-39fe2a01b248">
<div><br>
</div>
<div>Brainspace was originally targeted to the "legal discovery" problem, wherein you get a massive mailspool and have to search within it for things relating to your case. To do this, they have both the indexer and communication's path visualization tools
you would expect, but also a complete implementation of a arbitrary natural language parser for unstructured text and all the machine learning bits. So you could look for communications about, say, a nuclear facility at Natanz, where people didn't mention
nuclear or Natanz, in a set of documents that ranged in the millions, in whatever language you want. You can train it by saying "Yes/No" this doc is about bad things (tm), and it will bring you more of what you want to see. So that's....useful.
<span>😉</span></div>
<div><span><br>
</span></div>
<div><span>-dave</span></div>
<div><br>
</div>
</div>
</body>
</html>