# Do we provide a confidence score for our IP to Privacy data?

**URL:** <https://community.ipinfo.io/t/do-we-provide-a-confidence-score-for-our-ip-to-privacy-data/1723>\
**Category:** Knowledgebase\
**Tags:** database, ip-privacy-detection, extended-database, privacy-extended\
**Created:** [July 20, 2023, 5:13pm UTC](https://community.ipinfo.io/t/do-we-provide-a-confidence-score-for-our-ip-to-privacy-data/1723 "2023-07-20T17:13:17Z")\
**Posts on this page:** 1\
**Showing post:** 1

<div class="post-metadata">

**Author:** ![Abdullah](https://yyz1.discourse-cdn.com/flex035/user_avatar/community.ipinfo.io/abdullah/32/4973_2.png) [@Abdullah](https://community.ipinfo.io/u/Abdullah)\
**Post date:** [July 20, 2023, 5:13pm UTC](https://community.ipinfo.io/t/do-we-provide-a-confidence-score-for-our-ip-to-privacy-data/1723/1 "2023-07-20T17:13:17Z")

</div>

# We provide detection methodology information through our [IP to Privacy Detection Extended database](https://ipinfo.io/developers/privacy-detection-extended) that can be used to build an accuracy score

 ![building your IP privacy score](https://canada1.discourse-cdn.com/flex035/uploads/ipinfo/original/2X/0/0c605af44268222ab3c887d2479a7c2a624139ae.png)

We offer the [Privacy Extended database](https://github.com/ipinfo/sample-database/tree/main/Privacy%20Detection%20Extended) that opens up the methodology of our anonymous IP detection method. However, we don’t provide an accuracy metric as we do for our [IP to Geolocation extended database](https://github.com/ipinfo/sample-database/tree/main/IP%20Geolocation%20Extended).

The reason for this is that anonymous IP detection involves using multiple methods. We urge the user to understand the nuances of each detection method so that they can fine-tune the anonymous IP detection based on their cybersecurity models and needs.

If you have a malicious IP data log, you can write a regression/statistical model based on our dataset(s). You can even incorporate malicious IP logs, [IPinfo geolocation data](https://ipinfo.io/products/ip-geolocation-api), [ASN](https://ipinfo.io/products/asn-api) or [Company data](https://ipinfo.io/products/ip-company-api), abuse data, DNSBL data, etc., to create your privacy detection accuracy score.

[🔗 Documentation for IP to Privacy Extended Database](https://github.com/ipinfo/sample-database/tree/main/IP%20Geolocation%20Extended)

> **[Documentation for IP Privacy Detection Extended Database](https://ipinfo.io/developers/privacy-detection-extended)**
>
> We're the trusted source for IP address information, handling 50 billion IP geolocation API requests per month for over 1,000 businesses and 100,000+ developers

## IPinfo IP to Privacy Detection Methodology

Let’s explore the different data methodologies we employ in detecting an anonymous IP address. Each methodology comes with its own nuances and should be considered carefully if you are trying to build an accuracy score metric.

| Detection Methodology | Explanation | Caveat |
| --- | --- | --- |
| **anycast** | If IP is identified as being any anycast IP, that could map to multiple physical servers in different locations | Reasonable indicator. Not all anycast IPs are linked with IP privacy solutions. |
| **census** | If we’ve identified VPN software running on this IP as part of our internet-wide scan (successful openvpn or ipsec handshake) | Strong indicator. |
| **device\_activity** | If we’ve seen VPN-like behavior (multiple devices, multiple locations etc.) | Reasonable indicator. If a device is using organization/company VPN software, we might recognize that IP as a VPN. |
| **whois** | If we’ve seen VPN provider attributes in the IP whois data (e.g. provider name) | Weak indicator. As IP information in the WHOIS records tends to be comparatively static. [AS209854 Cyberzone S.A. details - IPinfo.io](https://ipinfo.io/AS209854#block-whois). |
| **vpn\_config** | If we’ve identified this IP in a VPN config file | Strong indicator. |

## Why we don’t provide an accuracy score

The reason we do not provide a single-digit indicator for our Privacy extended database is that it limits the potential for detection modeling of IP addresses. The more data points you have, the more effective your IP labeling models will be. You might have policies to accept some VPN traffic or some hosting traffic. Therefore, the custom modeling approach helps you fine-tune and personalize your access policies based on our multi-dimensional data.

Moreover, if you build your own IP privacy confidence score system, you can use our [IP to Geolocation](https://ipinfo.io/products/ip-geolocation-database), [ASN](https://ipinfo.io/products/asn-database), and [Company](https://ipinfo.io/products/ip-company-database) database to create a more robust cybersecurity threat detection policy.

* * *

### IPinfo services referenced:

- [IP to Privacy Detection Extended Database](https://ipinfo.io/developers/privacy-detection-extended)
- [IP to Privacy Detection API](https://ipinfo.io/products/proxy-vpn-detection-api)
- IP to Privacy Detection Database

Request your IP to Privacy Detection Extended Database by contacting our [sales team](https://ipinfo.io/contact).

> **[Contact Sales](https://ipinfo.io/contact)**
>
> Need help setting up our free IP geolocation API? Have feedback on the website? Get in touch

---

_[View the full topic](https://community.ipinfo.io/t/do-we-provide-a-confidence-score-for-our-ip-to-privacy-data/1723)._