This blog post explores some of the timeline and chart visualization capabilities you’ll need to consider when building a blockchain analytics tool.
But first, what is the blockchain?
In its simplest form, the blockchain is a shared digital ledger – a storehouse of transaction information. The ledger is decentralized and distributed, which means that each node on the network stores and maintains its own copy. New and existing data added to the ledger is validated using a unique text string (called a cryptographic hash), making it immutable and tamper-proof.
It is a novel and highly secure way of storing and sharing data. Its most popular application so far is as the engine behind cryptocurrencies such as Bitcoin and Ethereum, but it has much broader use cases ranging from supply chain logistics to medical records management, digital marketing, and real estate. estate.
There is a common misconception that because the blockchain is transparent (everyone on the network can see every transaction) and secure, it is easy to investigate. That is not the case.
Untangling the data from the blockchain to understand the underlying activity is a mammoth task. As the global market for blockchain technology grows, so does our need to manage and make sense of what happens to millions of digital assets.
In this post, we’ll focus on the cryptocurrency use case, but the concepts apply to any scenario where a person needs to interpret complex blockchain data to make an informed decision.
Why Use A Blockchain Analysis Tool?
According to coinmarketcap.com, the global value of cryptocurrencies ranges from 2 to 3 trillion dollars. And wherever you find currency, you’ll find criminals trying to exploit vulnerabilities.
Blockchain analytics tools are sophisticated platforms that help analysts connect actors and transactions with real-world identities, often to understand suspicious activity.
Governments, banks, virtual asset service providers (such as cryptocurrency exchanges) need a way to discover the people behind the alphanumeric strings. You can read more about who needs to analyze activity on the chain, and why, here.
The Three Steps of Blockchain Analysis
Before meaningful analysis can be performed, analysts must understand the entities and transactions in their data. This generally occurs in three phases.
The first is address classification. This is intended to connect pseudonymous blockchain addresses with real-world entities. Raw blockchain data is enriched with information from other sources, such as web scraping or dust attacks, and algorithmically clustered to group addresses most likely to be associated with a single controlling entity.
The transaction risk score then assesses the connections between the entities. It uses machine learning to evaluate each individual transaction on the blockchain and assign it a risk score, based on factors such as origin, wallet history, and money flow.
The third step is research. Here, analysts use sophisticated visual tools to dig deeper into their newly enriched data. This is where the visualization of graphs plays a fundamental role.
Graph Visualization For Your Blockchain Analytics Tool
Blockchain data is big, fast, complex, and full of connections. Investigating it requires specific graphical visualization and visual timeline analysis ability.
In no particular order, here’s our rundown of the 9 bits of functionality you’ll need to provide.
1. Integrated Time and Graph Views
The key to any blockchain data investigation is understanding what happened and when.
As a starting point, you’ll need to provide a chart and timeline view. Integrated together, the investigator sees the entities involved in each transaction and the order in which they occurred. Filtering on one view updates the other, so the researcher can explore activity across multiple dimensions simultaneously.
2. Flexible Filtering
On an average day, the Bitcoin blockchain facilitates over 250,000 transactions, in blocks of around 4,000. The data involved in blockchain analysis is large and growing rapidly. There are many techniques for reducing large graph data to a useful scale, but one key is filtering.
Researchers should be able to filter the data based on any attribute they want, from transaction size, wallet ID, risk score, or by time.
3. Insightful Designs and Smart Node Sizing
Even a leaked blockchain dataset is likely to be huge. Automated layouts are another powerful way researchers can start untangling large and complex blockchain data.
Some essentials are a powerful force-driven layout to display the entire network and a sequential view to untangle long chains of transactions.
Incorporating node size, either graph analysis measures such as degree centrality or attributes of individual nodes such as wallet balance, helps the researcher identify the people who play the most important role in a block chain network.
4. Easy Grouping
It’s Cryptocurrency best practice is to create new addresses for each transaction and spread the funds across multiple wallets. These precautions help ensure anonymity and reduce the risk of stolen funds, but they also frustrate investigators.
Address rankings go a long way in ordering the dataset, but they are never 100% accurate. Investigators need a way to manually combine different addresses and dive into the automatically combined ones as investigations progress.
5. Custom Styling
UX graphics display is its own issue, but a fundamental principle is to avoid overwhelming the user. Show them only what they need to see, but give them the tools to explore at their own pace.
Using techniques like tooltips, link styles, and donuts, you can add risk scores, wallet balances, transaction amounts, compliance indicators, and any other attributes to the chart, while maintaining a clean and clear user experience.
Time series charts help researchers identify events such as large-scale cashouts or transfers, or add additional contextual information, such as the bitcoin price, to show the rise and fall of the market.
6. User-driven Exploration
If we are putting the researcher in the driver’s seat, we must help him drive.
A common research approach is to start with a small data sample, perhaps just a blockchain address, and incrementally incorporate more data to build a bigger picture.
This ‘land and expand’ approach requires a lot of chart and timeline display features. Adaptive designs to add new data in a useful way to make it happen. An undo/redo stack provides easy reversibility when the investigator encounters a dead end. There is also node selection or entity pinning to help isolate specific activity in the chain.
7. Smooth Animation Animation
is an essential but easily forgotten part of a good blockchain analytics tool. With every layout, expansion, filter, or zoom in the timeline, there is a risk that the researcher will lose their train of thought or point of interest. Fluid animation keeps them on track.
8. Easy Dissemination of Intelligence
Intelligence is useless if it is not shared. Whether it’s a law enforcement officer investigating darknet trading, a VASP performing AML checks, or a financial institution filing a suspicious activity report, the investigator will want to share their findings with someone.
The ability to export charts, either as image files or custom PDF reports, is a key part of your workflow.
9. Powerful Graphics Rendering
Big data visualizations require great graphics rendering power. Investigators don’t want to stare at a loading screen for too long (or watch their display fail).
For your blockchain analysis tool to be successful, you need a high-performance renderer, such as WebGL. Other simpler approaches like SVG just don’t work as the data grows.