Visualizing SMS hotspots in days following the earthquake in Palantir.

Screenshot courtesy of Palantir Technologies

[Editor's Note: an edited version of this post first appeared on O'Reilly's Radar blog.]

The prologue was an earthquake of unexpected magnitude and location that left 250,000 dead.

As computer scientists and technologists, we’re used to dealing with large numbers in the abstract. Expressed in human terms, the mind-boggling numbers of 250,000 dead, 300,000 injured and over 1 million people left homeless are hard to comprehend.

Hit the link to read more about how effective data management and analysis is crucial to recovery efforts and see specific examples of data about the situation in Haiti modeled in Palantir Government.
Read the rest of this entry »

Software engineering is a craft that blends science and art. This fact is easy to overlook as the artistic aspects are often eclipsed by discussions of the science and technology behind what we do.

This is not one of those times: the art in software engineering is most evident when building compelling visual interfaces, something Palantir knows a thing or two about.

A demo reel is an industry term in the movie business — a short reel that acts as a portfolio when applying for jobs, a highlight reel of the author’s visual career. We’re not in the movie business, we’re in the software business. We do, however, use moving pictures to tell stories, stories backed by data — this is our demo reel: two-and-a-half minutes of data visualization and user interface eye-candy (It has pounding music — you may want to put on headphones or turn down your speakers.):

The movie will take a few seconds to load. It’s 800×600, so expanding to full-screen is suggested. We’ve done our best to create a streamable-yet-good-looking video. The compression artifacts are there, but shouldn’t be too distracting. In a real Palantir client, there are no compression artifacts and everything looks even better than it does here.

The Palantir family of products is much more that just pretty pictures; we have the underlying intelligence infrastructure to make those realtime animations possible and (more importantly) meaningful. That said, we sure do think they’re pretty.

By the way, if you’re interested in the progression of our interfaces, this not the first time we’ve posted eye candy: we posted a set of updated screenshots a little over a year ago; think of this as the next installment in the series.

And yes, it’s really all Java Swing.

In a previous post, Eric W. covered how we analyze polled system health information. Now we’ll look at pushed information, in the form of logging events.

Use Cases & Constraints

We decided on three kinds of questions we wanted to answer:

• What is the health of the deployment?
  • Example: What errors have occurred in the last 24 hours?
• Which parts of the platform are our users engaged with?
  • Example: How much time do users spend in each application?
• How is our server performing over time?
  • Example: What is the average wait on a search query?

The chief constraint was that we build our platform on Log4J. We already use Log4J all over the project, so converting the logging was out of the question. Besides, Log4J provides a guideline for the kind of metadata our events should support, and Log4J makes it easy to record events to a database.

That left us with two problems to solve: how to store structured data with a Log4j message, and how to analyze the collected data.

Analysis is the easy part: just use Palantir! After all, a sequence of logging events has a lot in common with a time series. The rest is explained below.

Read the rest of this entry »

We put up a post last year on the 2008 VAST Grand Challenge. Well, the IEEE VAST Challenge 2009 is over and the awards are in. We had another strong year, scoring two awards:

• Grand Challenge: Analyst’s Tool Choice (Of 48 submissions, only 3 Grand Challenge awards were given)
• Intuitive Traffic Visualization and Video Description of the Analysis Process

Some background on the event: three years ago, the IEEE began an annual conference called VAST (Visual Analytics in Science and Technology). The VAST symposium focuses on the fundamental research contributions and real-world application of visual analytics. As a part of the conference, the VAST Challenge allows teams to compete on delivering analytic solutions against a synthetic real-world dataset.

A selection of choice quotes from the judges:

• An award for “highly usable integrated exploration environment”, “efficient analytic exploration platform” or something along these lines would be appropriate.
• Survey Question: How much novelty do you see in this submission (data processing, visualization, interaction, hypothesis generation or evaluation, overall process, etc.)? Answer: More so than novelty was the extremely efficient solution approach to this challenge, much more so than other solutions.
• The submission shows two things very clearly: One, it shows the analytical process as being a multi-faceted, simultaneous processing of different information that is quite common among analysts. Two, it shows how multiple perspectives can be displayed on a single monitor, enabling the analyst to visualize what his mind is analyzing. Outstanding!

Our submission

And finally, our submission to the Grand Challenge. Here we have our overview video, with a link to the full video below:

For an in-depth look at the data and techniques used to make this a reality, check out our full submission in Finding a Mole: Cyber Counter Intelligence on the Palantir Analysis Blog.

I used to think I understood MVC. In undergraduate CS programs, MVC is taught as an off-the-shelf pattern, explained once and then ready for use in the real world. Wikipedia also makes it seem pretty simple:

Model–View–Controller (MVC) is an architectural pattern used in software engineering. Successful use of the pattern isolates business logic from user interface considerations, resulting in an application where it is easier to modify either the visual appearance of the application or the underlying business rules without affecting the other. In MVC, the model represents the information (the data) of the application; the view corresponds to elements of the user interface such as text, checkbox items, and so forth; and the controller manages the communication of data and the business rules used to manipulate the data to and from the model.

They go on to show the classic triangle diagram and how it’s baked into various GUI and web frameworks. There’s only one clause in the entire article that hints at something deeper: “Though MVC comes in different flavors…”

Different flavors indeed. In fact MVC is not just a pattern but a whole family of patterns: MVC, MVA, MVP, PAC, Model-Delegate…. It very quickly gets very hairy.

In this article I want to describe one of MVC’s lesser-known variants, the Model-View-Adapter (MVA) pattern, and talk about its advantages over traditional MVC in the context of a Java Swing application.

Read the rest of this entry »

As we mentioned in an earlier post, Palantir was recently invited to the IEEE’s VisWeek in Dayton Ohio, and was honored to be invited to participate in the VAST Interactive Challenge as part of VisWeek.

After winning an award for Interactive Visual Analytic Environment, Palantir was one of three teams selected to participate in the interactive session from 73 VAST Challenge entries. For the challenge, we were given a completely new set of data to analyze. We had 30 minutes to import 3 disparate datasets into Palantir, 30 minutes to train an analyst that had never used Palantir, and then 2 hours for the analyst to explore the data.

The data for the challenge came from three different sources, with a set of questions to answer for each set of data. There was an infectious outbreak, a Wikipedia edit war, and an abduction from a city park. Over the three challenges, there were over 100,000 datapoints to analyze. All of the data revolved around a fictitious town in Florida, Barracuda Springs, and was linked to the fictitious cult that was the center of the 2008 VAST Challenge. While two members of our team were importing the three datasets, the third team member was working with our analyst (each of the three teams was given a analyst from a nearby analytical organization). In 30 minutes, our analyst was able to learn how to conduct relational, temporal, geospatial, and statistical analysis in Palantir. After the 30 minutes of training, she was able to easily navigate the Palantir workspace, and solve all three challenges. Below is her work (hit the link to check it out).

Her conclusion was that Palantir was “viciously good software” and that she would be asking her boss if they could acquire Palantir for their work. Hit the link below to see screenshots and explanations for one of the challenge workflows.

We really enjoyed the VAST Challenge, and our experience at VisWeek. There were a lot of outstanding papers, posters, and speakers at VisWeek, and we were inspired by many fantastic visualizations that might soon make their way into Palantir’s Finance and Government Platforms. We are also looking forward to the 2009 VAST Challenge!
Read the rest of this entry »

I recently had the opportunity to attend a talk by Ben Shneiderman, a big name in HCI and professor at the University of Maryland. He showed off a bunch of really cool visualizations he’s invented over the years, aimed at advancing the field of data analysis. One of the visualizations, the Rank-by-Feature framework, looked immediately useful to our product. It’s an overview of multidimensional data that uses coloring in a matrix of correlations. I decided to add it into Palantir Finance as a proof-of-concept.

We have a scatter plot in Palantir Finance, but it’s not designed to compare your data across many variables (called metrics here). I extended it by adding a small triangular matrix control that we call the scatter plot Quick Select. The control gives you a visual overview of the data, and allows you to identify interesting metric pairs and then drill down into a scatter plot for any particular pair.

Assume I want to view a set of 10 metrics (shown above). With our regular scatter plot, you have to choose two of these metrics and plot them to see their correlation. If you want to know how every possible pair of metrics correlate, you have to manually perform an O(n²) input operation.

With Quick Select, you enter the metrics once and the triangular matrix is formed. Each square in the matrix represents a pair of metrics, and each square’s color is the correlation between the pairs. For example, green is a strong positive correlation, light green a less positive correlation, white is no correlation, and red is a strong anti-correlation.

In the highlighted square above, we’re comparing percent return and correlation with the S&P 500 over the past year. The red shows a strong negative correlation. This makes sense in light of recent market behavior: the S&P 500 has not done well over the past year, so companies that were correlated with it also performed poorly, while companies that moved in the opposite direction performed well.

Drilling down to a scatter plot is as simple as clicking on the corresponding square. Below are 2 of the 45 possible scatter plots defined by this set of 10 metrics. You can quickly jump back and forth between different scatter plots while retaining a nice, condensed overview of the data in the triangular matrix on the left.

This prototype took less than a day to write. And it was written from outside the system, using only the pluggability points Palantir Finance provides. Overall it’s a powerful visualization component, added as an extension to the Palantir Finance platform, and done in under a day. Pretty cool!

Many thanks to Prof. Shneiderman for the idea.

Carl juggles with his creation

One of the things our customers love to do is print our beautiful object graphs and tape them to the wall for discussion. What they hate to do is print 30 pages, line them up, and tape them to a poster one at a time. So we bought a plotter, and I started plotting.

I needed to print directly to a Java Graphics object. Unfortunately, the available information on large output printing from Java is thin at best. While there are lots of ways to successfully place ink on paper, I was only able to find one that reliably lets the application pick odd paper sizes that plotters use, like 24×19.7 inches. (The term “plotter” used to mean something with pens for printing blueprints and such. Now it just means a large format printer, commonly printers that can use roll paper as a source.)

One of the first things you’ll learn when you start working with printing in Java is that a language intended to be all things to all people (i.e., cross-platform) is utterly lousy at tasks highly specific to a given environment, such as printing. It will not surprise you to hear that native print services on Windows are pretty different from those available on a Mac, which themselves are pretty different from the CUPS system common to Unix systems.

So, by and large, you are reduced to the least common denominator of printing. Part and parcel of this least common denominator is agreeing on what constitutes a piece of paper and sticking to it. This is fine for people thinking, “My paper is 8.5 inches wide by 11 inches tall.” It poses a bit of a problem for people with plotters who are thinking, “My paper is 24 inches wide by as many damned inches tall as I need.” Even relatively powerful programs like PhotoShop or GIMP don’t seem to support plotters well. I believe Photoshop works by specifying the exact paper size you want to use, but any technique in which the easiest solution for the user is to pull out a calculator does not meet with my approval.
Read the rest of this entry »

Palantir has entered the 2008 VAST Challenge. We present an in-depth look at one of our challenge solutions as the first public example of the Palantir platform in action.

Two years ago, the IEEE began an annual conference called VAST (Visual Analytics in Science and Technology). The VAST symposium focuses on the fundamental research contributions and real-world application of visual analytics. As a part of the conference, the VAST Challenge allows teams to compete on delivering analytic solutions against a synthetic real-world dataset.

Each year, the organizers build a very vast (pun intended) dataset from scratch. The data is entirely fictional but mirrors real-world use cases and scenarios. This year’s dataset is about a new religious movement that started on an imaginary Caribbean island (cleverly titled Isla del Sueño, or “Island of Dreams”) situated between Florida and Cuba. There are four subsets of the synthetic data: the Wikipedia page for the movement and its associated edit and discussion pages; landing and Coast Guard interdiction records for boats leaving Isla del Sueño for Florida/Mexico; cell phone records from the island; and RFID tracking data from people in a building that was attacked with an IED.

The types of questions asked in the problem sets are qualitative questions that require answers backed by data. These are the sorts of questions that don’t yield answers using a machine-learning/data-mining approach nor can an unassisted human get these answers by simple inspection of the data. They require some sort of human-computer symbiosis to solve.

To solve the VAST problems, we assembled an ad-hoc team of analysts — composed of a mix of engineers, in-house professional analysts, and one senior executive — and asked them to use the Palantir Government software to extract insights from the data.

The results speak for themselves: the complete set of Palantir’s VAST solutions are available here.

Read on for an in-depth look at how we deconstructed and solved one of the problems.

Read the rest of this entry »

About 10 months ago, we released set of nine screenshots from our applications. Time has passed and we have not stopped working; the look of the applications has evolved. Here are some updated screenshots:

 
 

Read the rest of this entry »