> For the complete documentation index, see [llms.txt](https://docs.microscopeprotocol.xyz/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.microscopeprotocol.xyz/onboarding/data/source.md).

# Source

{% hint style="info" %}
Please note that while the information presented on this page is intended for **human consumption** and is accurate to the best of our knowledge, the official **source of truth** for our project is the corresponding codified files hosted on GitHub. We strive to maintain up-to-date and accurate information on this page, but please refer to the GitHub repository for the most current and reliable information.
{% endhint %}

> Github: <https://github.com/openchainmeta/chainmetareader/blob/main/chainmeta_reader/config/sources.json>

## Description

The ‘Source’ field provides information about where are the address labels (i.e. the address name and address entity name) from. The list of values available for this fields are provided in the table below:

<table><thead><tr><th width="149.33333333333331" align="center">Value</th><th>Description</th><th>Example</th></tr></thead><tbody><tr><td align="center">Ground Truth</td><td><p>Generated from: </p><p>1) direct interactions (transactions) with the labelled entity, or </p><p>2) official sources like OFAC or On-chain protocol page (e.g. Uniswap documentations)</p></td><td>Initiated deposit and withdrawals transactions against Binance to identify Binance addresses on chain.</td></tr><tr><td align="center">External</td><td>Generated from external parties that are available to general public - e.g like Twitter, Discord, Etherscan, etc</td><td>Binance PoR addresses from Binance’s official website</td></tr><tr><td align="center">Heuristics</td><td>Generated from rule based approaches that are used to clusterize addresses on chain based on their transaction patterns - e.g. common-spend, core addresses, change, etc</td><td>Common-spend clusterizers running on Airflow for UTXO based chain - e.g. Bitcoin, Dogecoin</td></tr><tr><td align="center">Machine Learning</td><td>Generated from machine learning based models</td><td>Based on ground-truth labels, we using GCN, XGB models to train and predict label for new addresses.</td></tr></tbody></table>

<br>