Guided Analytics: Paths to Threats in Big Data

 

INFORMATION OVERLOAD

In the age of big data, authorities tasked with border security have too much information to deal with, yet are responsible for quickly finding threats before they affect their country. Trained, experienced analysts are essential yet scarce resources. Training a new analyst is a process that spans months and years, and that requires significant investment on the part of the border security authority.

The answer to this has been to provide the analysts with ever more complex tools. These analytic systems provide the equivalent of a “workshop, materials, and tools” for a trained, seasoned, and practicing analyst who can then “build anything they want” from the available data.

These systems are consistently overwhelming for a new analyst, as they present huge amounts of data and palettes of analytics to choose from, usually with little supporting training or tutorials. Even for a trained analyst, knowing which analytics to use to answer a question is daunting – and leads to analysis paralysis.

We need a way to provide all analysts with a guided path to their objective while enabling advanced analysts to operate unguided based on their developed tradecraft.

GUIDED ANALYTICS

Systems developed with Guided Analytics focus on helping us get to our answers by giving us a clear set of starting points and then steering us toward our objective. We start with our objective, and the system helps choose the best tools for the job. To develop Guided Analytics, we must follow these principles:

Principles of Guided Analytics

Start with the objective, not the data

Set realistic limits and expectations

Automate a common path

Provide an off-ramp

Start with the objective, not the data

We interact with a system with a purpose in mind. It’s easy to lose that purpose when confronted with data – do we start with the traveler that matches our narcotrafficking profile, or the organization that they may belong to, or the address of the hotel they stayed at while in country? Guided Analytics lets us indicate our objective and then suggests data that helps meet that objective, rather than overloading us with hundreds of data points to connect.

Set realistic limits and expectations

Guided Analysis isn’t going to magically produce an answer in all cases. For a Guided Analytic to work, we must understand what it can and can’t do with the information available. For example, if we’re trying to establish a connection between travelers, a jihadist organization and prison radicalization, we will logically need intermediate data such as prison visitation records so that correlations can be made.

Automate a common path

Experienced analysts will develop tradecraft – ways of reliably getting from their question to an answer. We need to identify one of those paths to automate as a Guided Analytic. The best path is one which reliably produces an answer with data that we are likely to have. It’s best to reduce the number of options or customizations available to prevent confusion.

Provide an off -ramp

An experienced user may discover their own path to the objective, or even identify a completely new objective, while being guided. The system should provide a way to pursue this new objective, and if possible mark the point of departure so that the user can return to the guided path.

SECURIPORT’S ANALYTICS SUITE

Securiport’s analytics products were built on the principle of guided analytics.

For example, our Advance Passenger Information analytics suite provides tools designed for a minimally-trained user to investigate potential threats. An immigration authority can quickly react to an incoming person-of-interest in API data and meet their objective of investigating their travel history, companions, travel patterns, and global routes in a guided way. The system ensures that the user sees the important information first, without needing to navigate across pages of passengers and flights to nd the critical data. The system provides a step-wise common path of visualizations so that users aren’t overwhelmed with choices and are guided to next steps. A user, however, can choose to investigate the data in any order, or to change the focus of their investigation – and the system always provides a way back to any point in the investigation.

Even our advanced analyst portal – OctaneTM – was designed with the principles of guided analytics. With the ability to process, visualize, and explore millions of link-data entities, experienced analysts can build their case in limitless ways. However, OctaneTM provides guided analytics throughout the system, often in context of which entities are selected and which questions can be answered based on that selection. For example, “Shortest Path” analytics quickly answer the question “how are these two travelers connected?” or “Who else travels with these people?”

DETECTING AND STOPPING THREATS AT THE BORDER

Securiport’s Integrated Information Management System (IIMS) solutions provide threat detection, analysis, and decision support tools for of officials working across the border protection domain. From the immigration of officer to the threat expert to the intelligence analyst, Securiport’s tools enable layered defense across all areas of border security. This is The Science of Safer NationsTM.