Web

Intent Detection – Categories

Bolster’s Intent Detection model leverages the capabilities of  deep learning and natural language processing to detect the intent of a webpage. Currently, the model can classify webpages into the following categories.

Category
Name
Category Mnemonic (for API calls)DescriptionSample Webpage
Sensitive Datasensitive_dataWebpages asking for user’s login or personal informationSensitive Data Example
Shopping/
eCommerce
online_storeWebpages selling products/services onlineShopping Example
CryptocurrencycryptoWebpages with cryptocurrency-related contentCrypto Example
GamblinggamblingOnline gambling, casinos, and betting websitesGambling Example
GaminggamingOnline gaming websitesGaming Example
CaptchacaptchaWebpages asking users to fill in a captchaCaptcha Example
Promo Codepromo_codeWebpages claiming to give away promo codes of any brandPromo Code Example
Tech Supporttech_supportWebpages prompting users to call a tech support numberTech Support Example
SurveysurveyWebpages asking users to fill in a surveySurvey Example
Gift Cardgift_cardWebpages claiming to give away gift cards to usersGift Card Example
Hacked Websiteshacked_siteWebsites that have been hackedHacked Websites Example
Domain Parkingdomain_parkingWebsites that do not host active content, but are registered and parked for later use/ saleDomain Parking Example
Pharmacy/
Drug
pharmaWebpages containing content about pharmacies and drug storesPharmacy/Drug Example
StreamingstreamingWebsites streaming movies, tv shows, live games and other contentStreaming Example
Error Pageserror_pageWebpages that display error messagesError Pages Example
Directory Listingdirectory_listingWebpages that display the directory with files/ foldersDirectory Listing Example
BankingbankingWebsites related to banksBanking Pages Example
WarningwarningWebpages displaying a warning sign against potential phishing/ malicious pagesWarning Pages Example
HealthhealthWebsites that contain health-related contentHealth Pages Example
App storeapp_storeDistribution platforms for mobile applications onlineapp store Pages Example

Contact
contactWebpages offering users an opportunity to contact the organizationapp store Pages Example
BECbecWebpages that ask users to sign in, but are not associated with any brand. These webpages are used to carry out BEC scams.app store Pages Example
AdultadultWebsites hosting adult/ NSFW content