Email Validation

Various Methods Used for Validation

Common Mail Providers (including Free)

[https://github.com/derhuerst/email-providers/tree/master]
This NPM package download common mail providers. Loading this list, helps with skipping domain checks and more focuses on error in username.

Good domain Health == Good chances of not being Spam

[https://majestic.com/reports/majestic-million] Loading this into memory instance or redis or firebase can significantly improve domain check. We can enrich this data with time by verifying mx record, last updated, bounceRate Updating parameter like bounce rate can help by large number when trained on a huge data.


HIBP (Have I been Pawned)

[https://haveibeenpwned.com/] This website has more than 1 billion+ pawned email address. Most of these leaks are large and from reputed organisation who do thorough verification.

HIBP (Have I been Pawned) + Improvements

Since API is paid, during manual testing, this had better success with Common emails. We can also use this with rate-limiting to fully utilise API or even get such breached data by contacting or connecting research community.

OpenAI API

OpenAI prompts were very powerful in separating domain, tld and username. It can further analyse username and domain at deeper level and provide specific suggestions. (e.g., change gmial to gmail )

Python- MultiProcessing

For list, we can divide list into multiple parts and load in parallel. Since we will load popular domains, email provider and other details on memory - this is enhance process significantly


Disposable Domain List

[https://github.com/groundcat/disposable-email-domain-list/tree/master] As name suggest, this are temporary domain names and not great fit for subscriber building. While for paid service, this shouldn’t affect, but for free tools, reports and giveaway - we can enable this check and mark such email invalid

tld spam score

[https://www.spamhaus.org/statistics/checktld/pk] For each tld spamhaus keeps track of bad rating. If it’s >10%, those tld and less popular tld should be treated with caution

Built With

• email
• python
• spamhaus
• telegram-bot