The Intel­li­gence Community’s New Analy­sis Plat­form is the webi­nar I am attend­ing today. It is hosted by Carah­soft — the same peo­ple that hosted the Syman­tec Webi­nar I attended. The com­pany whose prod­ucts this is about is called Palan­tir Tech­nolo­gies. They have wide spread finan­cial and gov­ern­ment data analy­sis tools. Todays webi­nar focuses on the gov­ern­ment sectors.

From a side note while we are wait­ing for this to start Palan­tir, accord­ing to wikipedia, is an arti­fact from the Tolkien mythos. specif­i­cally: A palan­tír (some­times trans­lated as See­ing Stone but actu­ally mean­ing “Far­sighted” or “One that Sees from Afar”) is a stone that func­tions some­what like a crys­tal ball.

Webi­nar started

They start with the stan­dard thank you to their host carahsoft.

Palan­tir was started in 2004 by Alexan­der Carr (sp) and the pay­pal team. They pointed out that the main part of Pay­pal was anti fraud and that’s how they relate to the intel­li­gence com­mu­nity. They spent some­time dis­cussing Pay­pal and their com­peti­tors. The rea­son pay­pal suc­ceeded com­pared to their com­peti­tors was that they had ana­lysts to sort the data com­pared to using pure com­puter work. The ana­lyst then built tools to work with the data at the con­cep­tual level com­pared to the data level. This allowed the ana­lysts to help fight fraud bet­ter then the com­pe­ti­tion which allowed them to succeed.

Pay­pal then looked at what areas they could fit these types of tools. They rec­og­nized that they would fit best in the high finance and the intel­li­gence com­mu­nity. They worked and cre­ated tools and today’s web­cast was about the intel­li­gence com­mu­nity tools. Pla­natir is a front end and back­end tool

Data Inte­gra­tion — takes all your records and puts them into one uni­fied view that allows an ana­lyst to have an easy view of uni­fied data.

Search and Dis­cov­ery — see who the user com­mu­ni­cated with — per­sis­tent search which alerts the ana­lyst when new infor­ma­tion becomes available.

Link Analy­sis– the data­base includes his­tor­i­cal and auditible revi­sion analy­sis — I assume this is it help ensure the integrity of the data­base. Includes meta data for source, who added it, and where it came from, who updated it, essen­tially every step of change and data you could want. This also allows the revi­sion­ing data­base to look quickly at the data his­tory. They then go on to show dif­fer­ent views and his­tory track­ing they can utilize.

He shows how you could do dif­fer­ent infor­ma­tion extrac­tions to track ter­ror­ist activ­i­ties (once again using the ter­ror­ist threat to try to drive the point home /rolleyes). While this would be a great tool in gen­eral I think this could have been done another way instead of dri­ving the ter­ror­ist angle. The data set that it can pull from is quite large and very interesting.

Very inter­est­ing drag and drop inter­face for entity res­o­lu­tion and analy­sis. Very Web 2.0ish but yet seems they put too much empha­sis on the fisher price like inter­face. It’s not a bad thing just overly rounded — but I assume this may help things work faster — I have noth­ing to com­pare it to though. They are offer­ing a video at the end so I’m not going to describe all of the inter­face and his inter­ac­tion with it.

This does make Jim Cropcho’s dis­cov­ery of a flaw in the ohio vot­ing records a very triv­ial dis­cov­ery for this soft­ware and makes how we main­tain our records espe­cially our vot­ing and doc­u­ments that should be pri­vate (no your phone is not really con­sid­ered pri­vate sorry) oth­er­wise data dis­cov­ery with tools like this is trivial.

The plat­form does sup­port plug-ins.

When asked how many enti­ties they can han­dle they stated over 100 mil­lion enti­ties — they have pulled in all of IMDB, Wikipedia — and NETFLIX!!!! — so this com­pany is look­ing at user check­outs and rat­ings on net­flix — I’m going to fol­low up with net­flix and find out what of my pri­vate data is avail­able to these other companies.

More links

here
here
Flash Demo

For more carah­soft webinar’s go here for sign up.

  • Artificial
    he safest solution if you are worried about anyone tracking what you do online is to not go online. Otherwise someone will always be logging where you go and what you do, whether they look at the logs is another question
  • Actually it would in theory be mroe far reaching then that. You would have to avoid all video camera's, all credit card usage, no EZ pass, no cell phones, etc. If you don't want to be directly tracked online use a service like TOR. Just no it j anonymizes you, it doesn't necessarily give you privacy in the strict sense.
  • From the presentation I went to they used database download dumps of the wikipedia. From what I can gather (I haven't looked at the dumps that wikipedia offers myself) it includes edits, who edits them, where they are edited from, site content, and links between content types. I highly doubt all the log information about searches and who searched what is located in the databases since I would assume the logs are epically bigger then the wikipedia database itself.

    As for who gets to see the data. That may be a trickier question. They showed us some data set examples when I went to this webinar, but I would assume private government audiences get more of a true peek on how the inner workings go. So the government would have access to any type of data correlation that palantir can perform.

    Also they could use the netflix database, some search engines offer up their databases for research purposes, all of these things can be gathered and inputted into software such as palantir. The safest solution if you are worried about anyone tracking what you do online is to not go online. Otherwise someone will always be logging where you go and what you do, whether they look at the logs is another question, but software such as Palantir's make sifting that data trivial.
  • wruyg
    what exactly do they track on wikipedia? searches? what links every individual follows? who gets access to the info they're gathering?
blog comments powered by Disqus