• Y.O.R
  • Posts
  • Ethereum's Bot Ecosystem: A Machine Learning Expedition

Ethereum's Bot Ecosystem: A Machine Learning Expedition

Exploring the Diverse World of Automated Agents on the Blockchain

Theory & Proposal

The integration of bots in Distributed Ledger Technologies (DLTs) fosters efficiency and automation. However, their use is also associated with predatory trading and market manipulation

Nazari, M. H., & Biswas, A. K.

You know how crypto is all about decentralization and trustless interactions? Well, it turns out that the Ethereum blockchain, the backbone of the booming DeFi space, is swarming with bots – automated software programs executing transactions WITHOUT human oversight. These bots are like little digital workers, carrying out all kinds of financial operations on the blockchain.

But here's the interesting part: not all bots are created equal. Some are beneficial, like providing critical infrastructure or contributing to market stability. Others, however, are predatory, exploiting vulnerabilities or manipulating markets for profit. It's a digital Wild Wild West, with bots playing the role of cowboys, bandits, and pretty much everything in between.

Ai Dancing GIF by EsZ Giphy World

Gif by EcrookedletterZ on Giphy

Findings

We deploy both supervised and unsupervised machine learning models and measure their performance…

Nazari, M. H., & Biswas, A. K.

In a recent study, the researchers decided to take a closer look at this bot phenomenon. They combed through Ethereum transaction data, identified bot patterns, and even created a taxonomy categorizing different types of bots based on their behavior.

Using machine learning techniques, they developed models that can detect bots with an accuracy of up to 83%. They also found that certain features, like the timing and frequency of transactions, gas prices, and gas limits, are particularly useful in identifying bot activity.

But that's not all! The ended discovering that some bot categories, like those associated with centralized exchanges or decentralized trading strategies, can be clustered with high purity. It's like these bots have their own digital neighborhoods, hanging out together and doing their bot things.

Tax The Rich New York GIF by INTO ACTION

Gif by IntoAction on Giphy

Here are some details on the bot categories and taxonomy:

Subscribe to Keep Reading

Don't worry it's completely FREE!

Already a subscriber?Sign in.Not now

Reply

or to participate.