After writing a profile, uploading the best picture you can and sending countless witty opening messages, you get a reply from someone who is perfect.
You cannot believe your luck, but is this person too good to be true?
It’s a sad fact that most online dating sites have to combat fake / scammer profiles.
The romance scam draws people in every day and some of the high profile cases have cost innocent dating members up to $500,000.
The scam works like this:
- A criminal gang create fake profiles on dating sites.
- They build up an online relationship with their victim.
- They create a story to ask the victim for money.
The number of people who get caught by the scam is low, but with huge rewards the scammers have no intention of stopping.
A dating site that doesn’t fight hard to combat these scammers can soon become overrun with fake profiles, which ruins the experience for all members and ultimately destroys their business.
In the past 12 months at saucydates.com we are proud to say we have detected and deleted over 20,000 fake profiles, and add new technology continuously to improve detection rates.
To automatically detect a scammer, you need to compare their behaviour online to that of a genuine member, and track the differences.
We give all members 100 points and deduct points for each breach from normality.
When a member hits 0 points they are deleted.
User traits that are very common for a scammer result in a larger deduction of points than those that are not so clear-cut.
This scoring system is known on our site as the Trust System, and all members can visibly see each other’s scores and some information as to why they have lost points.
Some of the featured highlights of our Trust System are:
Image recognition: When a profile image is uploaded, it is checked against a database of previously banned accounts.
IP address: All IP addresses are checked to see if they are associated with a banned account. In addition, IP addresses are checked to see if they have been used by multiple accounts.
User reporting: Nothing beats human intervention, so members are encouraged to report fraudulent activity. We don’t hide the fact that scammers operate on our site, and lists of what to look for are shown on message pages.
Profile scans: Language and word patterns are analysed and scored accordingly. Scammers often use the same profile description multiple times.
Message ratios: Scammers have to send large quantities of messages as their success ratio is low. Tracking send to receive ratios gives an indication of irregular activity.
Captcha failures: Sending messages to other members in bulk will result in more failures of anti-spam codes that are entered when sending a message.
Encrypted URLS: A scammer will look at ways to automate message sending by writing their own HTML forms that inject code into a site’s own system. We encrypt all critical URLs in the same way as an online bank, offering greater protection. If a page is accessed with an incorrectly formatted URL, this triggers a reduction in points.
IP address country vs. claimed country: Checking the country selected by the user against the country determined by their IP address, allows you to identify potential fakes.
Local time vs. claimed country time: If a scammer fakes their IP address, or uses a proxy server, then JavaScript can be used to detect their computer’s local time. If their clock is set to GMT +6 and their selected city is GMT —5, this would be suspicious.
Additional methods: There are many more areas that we analyse, but in the interests of our own security, we will keep those confidential.
The final pioneering technique we use is a routine that makes associations between accounts.
If a scammer makes just one mistake on one account, then all of their associated accounts are automatically flagged and reviewed.
Accounts that need to be reviewed before deletion are managed via a mobile application, developed in-house.
This means that our team can delete an account, or multiple accounts, with a single button click from their phone or iPad at any time of day and in any location.
The extra work needed to develop this security is considerable, and more server power is needed to run anti-scam checks than the normal dating functions.
However detecting and deleting scammers is essential to providing the safest dating experience possible.
Hopefully in the future online dating will have to meet a minimum requirement for security and fraud prevention – making it difficult for scammers to operate within the dating sector.
Until then, anti-scammer technology will be implemented to varying degrees of success on the thousands of sites online today.
By Selcen King, Marketing Director at SaucyDates.com.
