Ever since the election, I have been writing, here, in comments and on my blog. I started my blog right after the election in November. I have not been real active on my blog of late, I have been deeply involved in some political activities in other venues.
Tonight, I was looking at the logs on my server and I ran across an entry for a visit and it had an address I could trace back so I decided to check this one entry out. It sent a chill down my spine until I remembered who we are dealing with (Janet Napolitano).
I visited the site at riverglass.com. Their description of their business model is thus:
RiverGlass is a software-based solution for government and business that finds and delivers information culled from expansive and diverse sources, wherever it exists — from internal communications via company Intranets and among diverse agencies, to news feeds, subscription services, websites and Internet user groups. Additionally, RiverGlass data analytics get smarter over time, refining search parameters according to positive results, thereby continuously delivering crucial, relevant data specific to your unique criteria.
Certainly Interested, I visited the “Products” page. Here, they clarify:
Governmental agencies and businesses face three major challenges in acquiring information: extensive, irrelevant information in search results; impossibility of knowing the best places to look for information; and an infinite need for new information and insights. RiverGlass technology continuously scours the Internet or enterprise intranet to discover new information relevant to your needs:
In other words they find and create interesting threads for government analysts to ponder. Further down the page, they expound:
Threat Mitigation: monitoring activities of groups intending harm to corporate or public assets and keeping abreast of socio-political happenings in locales of strategic importance
How do they deliver value? By assisting analysts in assessing the meaning of the text.
…any particular analyst or knowledge worker has not just one but perhaps hundreds of persistent information needs and many of these information needs cannot be simply characterized by some small number of keywords.
Further, they have a proprietary method of statistically analyzing text to extract Points of meaning :
complex meaning extraction, the statistical approach has some limitations — it essentially treats natural language texts as sequences of words, but doesn’t have a knowledge model of what these words actually mean. These statistical tricks can be effective, but at their core they are tricks to approximate the extraction of meaning, and are subject to some fundamental limitations relating to polysemy (words can have multiple different word senses) and to the variety of ways in which natural language can express meaning. Some hybrid approaches employ the syntax of sentences, which can provide some contextual clues, but syntax is not semantics — and there is no attempt to get at the actual meaning expressed in such natural language texts.
The thesis of OntoSem is that interpreting the meanings of textual units is really feasible only in the presence of a detailed world model whose elements are triggered (directly or indirectly, individually or in combinations) by the appearance in the input text of various textual units whose lexicon entries contain pointers to certain ontological concepts.
Enthralled now, I investigated further and found how they serve government specifically by spying on U.S. citizens:
The most critical issue facing intelligence operations is keeping pace with the continuous generation of new, disparate and obscure data. Tools that continuously analyze and automatically connect incoming data to open investigations are a necessity for intelligence-led security and policing.
It gets even better:
RiverGlass and the Statewide Terrorism Intelligence Center (STIC) of the Illinois State Police (those ones from Dear Leaders Homestate) have jointly developed a suite of tools tailored to the specific needs of analysts working in a fusion center environment.
In case the term “fusion center ” strikes a chord, let me refresh your memory, the Fusion Center in Missouri (MIAC) was the source of the infamous memo casting disgruntled ex-military, anti-abortion activists, and “anti-immigration activists” and angry white republicans in general as “potential” homegrown terrorists.
To encapsulate the purpose, in their own words:
Automatically seek out and monitor websites that relate to criminal investigations or other intel needs. Information posted to the internet from diverse sources like CNN, MSNBC, Southern Poverty Law Center, the Anti-Defamation League, and Department of Justice, as well as sites that post threatening or criminal content is critical to effective data analysis.
So, after a thoroughly entertaining hour reading up on them, I went back through all my logs in the last year. Yep, they visit every month. I have also had some visits from Centcom (Miami & Iraq) and visits from some other military posts (security gates) whom I assumed came to look at at a salutory post I made to all our warriors and to one specific fallen warrior who happened to be a friend of one of the regular contributors here.
So all the mea culpa’s from Janet Napolitano aside, DHS is still hot on the trail of me, and I sure a great many of you here at Redstate. I have only two hopes:
- That most of their math genius’s are not late of the Wall Street derivatives market.
- That their programmers are not late of Phil Jones office at HADCRU.
Oh, and they were not the only one, see this link .
In case you are wondering what is so interesting to them, see my blog: The41stVote .
I do have a sense of irony though, for you geeks out there, look at the source for this page and note the “alt text” tags I put on the links in here. That should give you a chuckle. I would also encourage all of you here that run a blog on you server to look for these guys. They will show up as and entry such as:
Cross posted at the41stvote.