From Digital Footprints to Digital Journeys, How AI-Powered Analytics Surfaces Insights
93% of investigators agreed that mapping and analyzing the journey of a subject is a top priority for any investigation – Source: 2019 Cellebrite Industry Trend Survey: Law Enforcement
With the expansion of personal data services, modern-day crimes increasingly include digital “footprints” of suspects and victims. However, even the trail of one person can include many digital channels, devices, and diverse data formats.
The challenge for law enforcement worldwide is to identify all relevant data points for an investigation in a timely manner. For digital forensics investigators familiar with this challenge, the inevitable conclusion is that the power of Artificial Intelligence (AI) and the flexibility of machine learning is necessary to deal with the onslaught of data that can hide essential evidence.
“It’s not information overload. It’s filter failure” – Clay Shirky, Consultant on the social and economic effects of Internet technologies
Cellebrite believes multiple and diverse digital footprints can successfully be combined with the assistance of Artificial Intelligence to form a broader suspect or victim’s journey embedded inside the data of an ongoing investigation.
The unique ability of AI to connect seemingly unrelated details and map significant activity can lead to a rapid output of insights that complement traditional investigative techniques. Until now, manual and labor-intensive activities have dominated this part of investigations. The time-consuming nature of piecing together scenarios of crimes and people involved through digital data has meant that many cases have gone cold or closed altogether.
The release of Analytics 8.0 changes the way agencies of all sizes can manage digital data in investigations.
The Analytics solution was developed to help investigators quickly know what to focus on, identify persons of interest, as well as pinpoint their actions across media, communication, and other event types. By also mapping locations, activity patterns and behavior anomalies it is possible to easily communicate the crime narrative to stakeholders.
Let’s explore Google services as an example of how AI is already making our lives easier in a and vastly change what we expect from data, media, and personal life management.
How We Consume Data has Changed with the Introduction of AI
Google photos is a great example of AI organizing media files in our private lives while adding new layers of value such as auto-created videos, image categorization, and people identification. The interplay of AI with personal pictures and videos can produce new media output on the fly, while tagging photos and maintaining searchable databases that are shareable and expandable on terabytes of storage space.
Google essentially leverages its access to millions of human-categorized picture and video databases to train its deep neural network. This supervised learning method compounds the accuracy of its algorithms to inform auto-suggestions and categorizations of the user’s media. All this computing power is accessible through the average smartphone with a stable internet connection and a cloud account.
As smartphones begin to add NPU (Neural Processing Units) to standard CPU processing, a new form of computational photography is being introduced that leads to smart hardware mechanics as well as instant person identification and photo categorization in real-time.
With Analytics 8.0, these types of features and benefits are offered with specific functionality for the digital forensics community.
Surfacing Key Insights to Focus the Investigation
Imagine, for a second, the following scenario: you have a high-profile case on your hands, a mountain of digital data, and somewhere inside there is that crucial piece of evidence you need to solve the case before time runs out.
All digital forensics practitioners know that to solve a crime, speed is of the essence. However, with the growing data storage capacities of smartphones and cloud accounts, the sheer volume of data from varied sources, in disparate formats, presents a daunting task for investigations running against the clock. Investigators need tools to help them delve deep into all relevant data details to get the answers they need quickly and reliably.
For these reasons and more, Cellebrite developed the AI-powered Analytics solution to surface insights right from the start without needing to filter, parse or categorize data manually. Analytics simplifies and automates complex analytical tasks and cultivates more leads in less time. Now, it’s possible to discover key evidence that can solve a case faster with cutting edge neural network-based machine learning algorithms.
Investigators who need to focus their investigation on what threads to pull in the first 48 hours are now provided with insights that surface seemingly unrelated events, locations and relationships to help zero-in on the digital evidence that matters.
Let’s explore the first thing you see when viewing the Analytics dashboard.
Right away, the Analytics dashboard displays relevant details including key insights that it calculated would be important for you to look at at the beginning of a digital investigations. It consolidates all of the data points found through media categorization, contact analysis, and core third-party interaction into a convenient lobby of decision that highlights the Who, What and Where.
This birds-eye view reveals connections between persons of interest while highlighting the patterns and anomalies in their behavior. Digital forensics officers can review aspects like the preferred methods of communications of a suspect, or victim, including their top interactions, places they visit frequently, and places they never visit until the time of the crime.
By Anastacia Ezrets - Cellebrite Analytics Product Manager