Risk assessment based on listing information

1. A system, comprising:

a non-transitory memory storing instructions; and

one or more hardware processors coupled to the non-transitory memory and configured to execute the instructions from the non-transitory memory to cause the system to perform operations comprising:

connecting to a listing webpage of a merchant site;

scraping listing data from the listing webpage;

transforming the scraped listing data into variables for a listing risk classifier;

inputting the variables into the listing risk classifier, the listing risk classifier trained using known fraudulent and legitimate listings;

determining a risk value from the listing risk classifier;

determining that the listing webpage is fraudulent based on the risk value and at least one predetermined threshold value for at least one of the variables;

receiving a user selection, wherein the user selection includes a confirmation label for the listing webpage that confirms the listing webpage as legitimate or fraudulent; and

training the listing risk classifier based on the risk value, the variables, and the confirmation label.

2. The system of claim 1 , wherein at least one of the variables is a ratio of words in uppercase to lowercase.

3. The system of claim 1 , wherein at least one of the variables is a number of a particular word appearing in the listing.

4. The system of claim 1 , wherein at least one of the variables is based on punctuation marks.

5. The system of claim 1 , wherein at least one of the variables is based on an identifier of a product in the listing.

6. The system of claim 1 , wherein the listing risk classifier is based on a random forest algorithm.

7. The system of claim 1 , wherein the operations further comprise preventing a purchase based on determining that the listing webpage is fraudulent.

8. A computer implemented method, comprising:

connecting to a listing webpage of a merchant site;

scraping listing data from the listing webpage;

transforming the scraped listing data into variables for a listing risk classifier;

inputting the variables into the listing risk classifier, the listing risk classifier trained using known fraudulent and legitimate listings;

determining a risk value from the listing risk classifier;

determining that the listing webpage is fraudulent based on the risk value and at least one predetermined threshold value for at least one of the variables;

receiving a user selection, wherein the user selection includes a confirmation label for the listing webpage that confirms the listing webpage as legitimate or fraudulent; and

training the listing risk classifier based on the risk value, the variables, and the confirmation label.

9. The computer implemented method of claim 8 , further comprising reporting the determination that the listing webpage is fraudulent to the merchant site.

10. The computer implemented method of claim 8 , further comprising denying a payment authorization based on determining that the listing webpage is fraudulent.

11. The computer implemented method of claim 9 , further comprising: receiving a user request for a listing risk analysis; and

reporting to a user device the determination that the listing webpage is fraudulent.

12. The computer implemented method of claim 8 , wherein the confirmation label confirms that the listing webpage as legitimate and the determination that the listing webpage is fraudulent as incorrect.

13. The computer implemented method of claim 12 , wherein the confirmation label confirms the listing webpage as legitimate, and wherein the method further comprises requesting additional proof that the listing webpage is legitimate.

14. A non-transitory computer-readable medium having stored thereon instructions executable by a computer to cause the computer to perform operations comprising:

connecting to a listing webpage of a merchant site;

scraping listing data from the listing webpage;

transforming the scraped listing data into variables for a listing risk classifier;

inputting the variables into the listing risk classifier, the listing risk classifier trained using known fraudulent and legitimate listings;

determining a risk value from the listing risk classifier;

determining that the listing webpage is fraudulent based on the risk value and at least one predetermined threshold value for at least one of the variables;

receiving a user selection, wherein the user selection includes a confirmation label for the listing webpage that confirms the listing webpage as legitimate or fraudulent; and

training the listing risk classifier based on the risk value, the variables, and the confirmation label.

15. The non-transitory computer-readable medium of claim 14 , wherein the operations further comprise delisting the listing webpage in response to the determining that the listing webpage is fraudulent.

16. The non-transitory computer-readable medium of claim 15 , wherein the confirmation label confirms the listing webpage as legitimate, and wherein the operations further comprise requesting, from a user device, additional proof that the listing webpage is legitimate.

17. The non-transitory computer-readable medium of claim 16 , wherein the operations further comprise relisting the listing webpage in response to receiving additional proof that the listing webpage is legitimate.

18. The non-transitory computer-readable medium of claim 17 , wherein the additional proof is an image.

19. The non-transitory computer-readable medium of claim 16 , wherein the operations further comprise updating the listing risk classifier in response to receiving the additional proof.

20. The system of claim 1 , wherein the user selection further includes an indication of a successful transaction associated with the listing webpage, and wherein the training the listing risk classifier is further based on the indication.