content filter

Content Filtering

Filtering with flexibility

Fast and accurate filtering in 18+ languages to best serve your business needs. Use one or more of the of the content filter options detailed below, all easily accessible via REST API.

intelligent chat filtering

Blacklist Filter

Blacklist filtering is used to identify inappropriate words and phrases in a collection of text. CleanSpeak uses Natural Language Processing and proprietary algorithms to determine if a text contains a blacklisted word or phrase.

Every blacklisted word CleanSpeak detects is classified with a tag (e.g., Sexual, PII, Vulgarity, Bullying, and many more) and a severity (e.g., Mild, Medium, High, Severe). You can decide to either handle the list of CleanSpeak detections yourself or instruct CleanSpeak to apply an action to the results. By applying moderation rules, CleanSpeak’s analysis results in a simple direction (e.g. allow, replace, reject, pending). Everything is configurable via a simple web interface making it easy to change the tolerance of the content filter without changing a single line of your application’s code.

Our approach has been developed over a decade and has been battle tested in a diverse set of applications: from games processing billions of chat messages per month to fortune 500 companies analyzing customer communications. We’ve got you covered.

blacklist filter screen shot

Keep Users From Bypassing Your Filter

To keep false positives to a negligible minimum, we focus on how users communicate to prevent them from modifying inappropriate text to sneak past a content filter. CleanSpeak catches these common tricks, among many others:

seperators filter
grawlix filter
leet speak filter
Leet Speak
phonetic replacement filter
Phonetic Replacements
repeat characters filter
Repeat Characters
embeddable word filter
Embeddable Words
whitelist filter

Kid's Chat Filter

Kid's Chat Filtering is the opposite of blacklist filtering and only allows acceptable words and phrases.  Anything that is not included in the whitelist will be rejected. Typically targeted at young audiences (under 12 years old), Kid's Chat Filtering is much stricter than blacklist filtering.  This allows for a safer environment. The sacrifice you will need to consider is that Kid's Chat Fltering can limit users’ ability to communicate which can then lead to frustration.

username filter

Username Filter

Usernames present a unique use case as they aren’t “typical” words that would be used in standard communication. Oftentimes words and numbers are included in a username, like Joe123. This can present a significant problem for a simple content filter, resulting in either too many false positives or false negatives, ultimately frustrating your users. CleanSpeak has a specific filter designed for usernames that implements advanced language analysis and more aggressive rules to ensure that only appropriate names are allowed.

email phone filter

Email & Phone Filter

You should not share personally identifiable information (PII) in public areas of the internet. Still, some people ignore this or, worse, can be potentially tricked into sharing their info for something other than a core competency of an app/website etc. CleanSpeak includes email and phone number filters for user protection.

url filter

URL Filter

In some cases you will want to limit user’s ability to include URLs in their communication in an effort to avoid spamming. CleanSpeak has an advanced URL filter to ensure they will not be shown.

The URL Filter Whitelist, similar to the Kid's Chat Filter, allows you to configure specific URLs or URL patterns that CleanSpeak should ignore when filtering. For example, you might want to configure CleanSpeak to ignore URLs inside your domain using this feature.

Ready to get started?

Sign up for your free trial today.