METHOD AND SYSTEM TO DETECT OUTLYING BEHAVIOR IN A NETWORK-BASED MARKETPLACE

• US 20100332346A1
• Filed: 08/27/2010
• Published: 12/30/2010
• Est. Priority Date: 04/12/2004
• Status: Active Grant

• Current Assignee: eBay Inc.
• Original Assignee: eBay Inc.
• Inventors: Embree, Kevin
• Granted Patent: US 8,260,681 B2
• Time in Patent Office: Days
• US Class Current: 705/26.35
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0282 Rating or review of business operators or products
• CPC Class Codes: G06Q 30/0609 Buyer or seller confidence or verification

Abstract

A system to detect outlying behavior in a network-based marketplace automatically collects attribute information for a first plurality of sellers that includes a first seller, and stores the attribute information in a storage device. The system computes peer information associated with a second plurality of sellers using a computer system, wherein the first plurality of sellers includes the second plurality of sellers, and wherein the peer information is automatically computed from the attribute information for the second plurality of sellers. The system compares the peer information associated with the second plurality of sellers with attribute information for the first seller for the purpose of automatically detecting outlying behavior by the first seller.

Claims

1. A computer-implemented system to detect fraudulent activity in a network-based marketplace, the computer-implemented system comprising:

a processor; and

a medium executed by the processor, the medium including;

a collection module to collect attribute information for a first plurality of sellers, wherein the collection module compiles a plurality of suspect sellers based on the attribute information for each seller in the first plurality of sellers;

a computing module to compute peer information associated with a second plurality of sellers, the second plurality of sellers comprising a subset of said first plurality of sellers, the peer information computed by combining together the attribute information of the second plurality of sellers;

a comparison module to compare the peer information associated with the second plurality of sellers with attribute information for each suspect seller in the plurality of suspect sellers; and

a detection module to detect outlying behavior by each suspect seller based on the comparison.

2. The computer-implemented system of claim 1, wherein the first plurality of sellers includes sellers that have listed an item for sale on the network-based marketplace.

3. The computer-implemented system of claim 1, wherein each seller in the second plurality of sellers is associated with a particular country such that the comparison module compares the attribute information for each suspect seller that is from the particular country with the peer information that is associated with the second plurality of sellers.

4. The computer-implemented system of claim 1, wherein the collection module compiles a plurality of suspect sellers by comparing attribute information for each seller in the first plurality of sellers for a current period with the same attribute information for at least one prior period.

5. The computer-implemented system of claim 4, wherein comparing attribute information for each seller in the first plurality of sellers for a current period with the same attribute information for at least one prior period includes comparing listings closed in a current week with listings closed in a prior week.

6. The computer-implemented system of claim 4, wherein comparing attribute information for each seller in the first plurality of sellers for a current period with the same attribute information for at least one prior period includes comparing gross merchandise sales in a current month with gross merchandise sales from a prior month.