The Palantir Government platform is a powerful environment for exploring the underlying connections within complex datasets. Revealing these connections is accomplished by separating the user from difficult and technical commands. Instead the user is presented with simple, yet powerful, interfaces allowing them to ask natural questions in an intuitive fashion. In order to perform the investigation called for by the Grand Challenge, we had to work with four sets of very different data. With Palantir’s data integration capabilities and Dynamic Ontology, we were able to easily bring all the VAST data together in one conceptual view.

Figure 1: The Data Import GUI

            Once the disparate data sources were loaded into Palantir and ‘published’, anyone with an account on our network could access and analyze the data from our webstartable client. Because the Grand Challenge has no data independent of the Mini Challenges, the Grand Challenge investigation required us to aggregate the analyses our team had already accomplished for the Mini Challenges. A detailed accounting of the processes used in the Mini Challenges can be found linked from our main VAST page. Below we will provide a brief summary of those Challenges and describe how they relate to the Grand Challenge.

            Wikipedia Analysis: Analysis of the Wikipedia edits and discussions was divided into two approaches: user-centric and time-centric. The user-centric team used Palantir’s Histogram to isolate the top discussors and editors of the Paraiso Manifesto page. The Histogram provides aggregate information and an overview about all selected objects in any of the Palantir Applications. Analysts could read the words of wiki contributors directly in Palantir, and were thus able to get a feel for the factions and major players participating in the discussion. This user-centric analysis was facilitated by the time-centric team, working in parallel, who used Palantir’s Timeline to dissect the flamewars (or areas of high activity). Our analysts noted rivalries between two wiki contributors and were able to isolate the pro-Paraiso wiki users in the discussion very quickly.

Figure 2: The “link by” option can quickly find related entities, events, etc.

            Migration Analysis: Boat migration data was divided into three major investigations: landings, interdictions, and launches. While our analysts could have analyzed all these aspects of the migrations within one investigation, breaking them down in this way helped to maintain a primary focus. The Analyses were enabled by Palantir’s integration with GoogleEarth to explore the geospatial variation in landings, launches, and interdictions over time. Our analysts noted interesting geographic zones in GoogleEarth (specifically, 5 distinct launch sites and 4 landing zones) which served as starting points for further investigation. It was simple to perform advanced analysis as well, such as identifying the number of ‘go fasts’ leaving from launch site 4 that successfully landed in West Florida in 2007. Interesting results were uncovered including: strategic weaknesses in Coast Guard interdiction patterns, a correlation between where boats launch and where they land, a temporal progression of launch sites, and a month-by-month chart of landing success rates.

Figure 3: All landings viewed in Google Earth

Figure 4: Palantir’s Timeline—times with Mexican landings are highlighted (yellow)

            Communication Network Analysis: To explore the communication network, our analysts began from the given starting point that Ferdinando Catalano is ID 200. Our analysts then added his immediate connections to the graph and subsequently second degree connections were added as well. These first and second order graphs of Catalano’s network were used to assign IDs to his inner network—Esteban Catalano, David Vidro, Juan Vidro, and Jorge Vidro. Analysis of 200-centric network using the Palantir Timeline revealed sudden network silence on the last three days of the ten day period represented by the data. This silence was not present in the overall network of some 400 IDs. This aroused suspicion that there was something duplicitous occurring in this inner network. Palantir was then used to identify the most prominent IDs during the period of network silence and uncovered a network that mirrored the 200-centric network observed earlier. This new network was centered on ID 300 possibly indicating that the inner network had switched cell phones. Additionally, the various movements of the telephone IDs were analyzed using the geographic position of the originating cell phone towers through visualization in GoogleEarth. Palantir’s ‘Intermediary Framework’ which allows phone calls to simultaneously act as links between people and independent events was very valuable in enabling this task.

Figure 5.1: The network of IDs 1, 2, 3, and 5, with an auto-layout in Palantir

Figure 5.2: The auto-layout of 306, 309, 360, and 397; strikingly similar to 8.1

            Evacuation Analysis: Because the data of the evacuation from the attacked Department of Health building is largely geospatial in nature, Palantir’s geospatial integration capabilities were heavily relied upon to analyze the attack. Importing a slightly transformed version of the evacuation data (with real geotime information) then exporting the results to KMZ from Palantir allowed our analysts to easily produce a time-lapsed animation of the evacuation. After adding an imaginary building constructed according to the given grid, the animation was closely watched in an attempt to identify who the potential bomber was, where the bomb went off, who could be a witness, who moved in a suspicious manner, etc.

Figure 6: After the bomb explodes

            With this foreground knowledge established, the next step was to explore the background of the movement. Understanding the context of a group is always critical to analyzing their activities. This is true in history when studying ancient civilizations, in sociology when performing demographic analysis, and especially in intelligence when conducting an investigation. It is, therefore, extremely important that we base our exploration of Catalano’s Paraiso Manifesto Movement (CPMM) on the Movement’s beliefs and its previous activities.

            Almost everything that can be learned about the Paraiso Movement must be gleaned from open source information, namely Wikipedia. By reading the actual page on the Manifesto and using Palantir to isolate important comments by editors of that page, it is possible to paint a reasonably detailed picture of what the Movement espouses (challenge one). First, followers consider the Movement to be a religion, not simply an organization committed to certain goals. In fact, Ferdinando Catalano claims to preach the true form of Christianity, though he does not respect the authority of existing Christian groups. As a corollary, believers view Catalano as a godlike, enlightened religious leader. Second, Paraiso proponents advocate a radically decentralized society in which the family-unit is supreme. The state exists to provide social services and infrastructure to families but should not intervene in disputes and has no coercive authority. In fact, comments in the Manifesto page’s edit logs indicate that the Movement has its own justice system and does not respect the convictions of other legal bodies. Third, the Manifesto declares that healthcare can only be approved by the head-of-household (father) and that even he is forbidden to approve antibiotics, vaccines, or other similar medicines. The synthesis of beliefs two and three produces a religious movement that is strongly opposes state-provided healthcare. Fourth, Paraisos consider women to be subordinate, confining their education to the home and prioritizing the desires of male family members ahead of their female counterparts. Finally, the Manifesto’s intellectual influences include “Pancho Villa, Che Guevara, Leon Trotsky, [and] Cuban revolutionary Jose Marti” a list of military commanders and revolutionaries with a history of violent actions. When combined with CPMM’s disregard for government-imposed laws, this suggests that the Movement may approve of violent tactics.

            Establishing this repository of beliefs gives provides an illustrative starting point for analyzing the activities of the Paraiso Movement. In observing the edit logs and verbal discussion of the Wikipedia page on the Paraiso Manifesto, we can observe a sense of persecution by many proponents and censorship of most criticism. Their self-proclaimed status as a religious group makes both of these likely, as religions have employed the rhetoric of external persecution for centuries and often prefer dogma to open discussion. Moreover, the fact that believers consider Catalano ‘enlightened’ could make any critical or derogatory comments seem legitimately offensive. The beliefs of proponents, therefore, explain the fervor with which they defend their public reputation on Wikipedia.

            There are also several pieces of evidence—including external suggestion and the presence of individuals with ties to high level operatives—that point to Paraiso involvement in the bombing of a Department of Health building located in Florida. An examination of their core beliefs makes involvement seem even more plausible (challenge four). We know that the Movement is unsupportive of healthcare, especially when it is government-provided. We also know that the family unit takes absolute primacy over the state and that there is little regard for external law. A wiki edit has been identified which implicates the movement in gunning down medical professionals. Finally, it is known that many of Fernando Catalano’s influences are violent. It is with these pieces of information that it was concluded that the Paraisos have both motive and likelihood to commit a violent action against a government health facility in one of their population centers (Florida).

            CPMM began in Catalano’s home country, Isla del Sueño, a Caribbean island between Cuba and Florida. The government of that island has increased pressure on the Movement, leading to mass migration (challenge two). This crackdown was probably precipitated by the organization’s previously examined violent undercurrent and disregard for the laws of the government. Moreover, its patriarchal demand that women be isolated probably produced insular and hard-to-control communities. Because Paraisos consider themselves religious in nature, and not simply political or social, it was likely that they would respond by migrating, as religions have in the face of oppression for centuries. Moreover, their emphasis on the family-unit explains the fact that many of the travelers on each boat shared last names. By the end of the three year period during which their migrations were tracked, emigrants from Isla del Sueño had landed in large numbers all along the coasts of Florida and Mexico’s Yucatan Peninsula, bringing the Paraiso religion with them (challenge two). We also know from the Manifesto’s Wikipedia page that smaller migrations and public advocacy (Pirate Radio) helped establish bases in many other Caribbean islands (challenge one). Because we have no tracking of movement on land in both Florida and the Yucatan Peninsula, it is also possible that the Movement has spread to other areas of America and Mexico, and possibly other Central American countries such as Guatemala, Belize, Honduras, and El Salvador.

            The next step was to decide specifically who is connected to the Paraiso Movement and what roles they might play. Members of the movement were identified slightly differently for each investigation (Mini Challenge). Among Wikipedia users, those who were identified as fervent defenders of the Manifesto were considered members of the network. The cell phone data had only five associated names, all of whom are high level operatives that clearly belonged to the network. Obviously, many of the numbers called by the Catalanos and Vidros are related to the movement, but there is insufficient information to disambiguate a number from a member from an unrelated third party. In the evacuation data, we added Ramon Katalanow and Carlos Vidro for their apparent familial ties to known important figures. We also added several suspicious individuals, including the suspected bomber, Cleveland Jimenez. The most challenging decision, however, was for the migration investigation. Nearly every person escaping from Isla del Sueño could be a Paraiso believer attempting to escape the government crackdown. However, most of the Plymouth Pilgrims weren’t Puritans; they were simply trying to take advantage of a chance to move, and a similar situation could be occurring here. Moreover, simply because the migrants believe in the religion, it is not possible to categorically consider them part of its social network if they are not actively connected to other major players. Therefore, only Eduardo Catalano and Jesus Vidro were added—once again for their familial connections. The players pulled from each investigation were tied to each other, but they were also tied to everyone else in the network through a presumed tie among Catalanos/Vidros (with the exception of the Wikipedia edits which had no obvious members of either family).

            The sole remaining task at this point was to figure out which members of the movement might be responsible for specific activities. Since individuals had been identified based on investigations related to entirely distinct activities, the work was mostly done already. Members drawn from the Wikipedia investigation are involved with public relations and information control. Important figures in the migrations are illegally immigrating into new countries to establish bases and spread the religion. High level operatives being tracked by cell phone are responsible for coordination and communication. Finally, those identified as involved with the DoH bombing are violently active agents. The range of activities that the Paraiso Movement participates in may be much broader, but those actions are outside the scope of the data and investigation. Given the information uncovered by Palantir, a continued operation could easily uncover much more detail about potentially violent cells of movement and combat the threat the Paraiso Movement poses as it spreads its reach.