Cursor path vector analysis for detecting click fraud

US8935176B2 · kind B2 · utility

Assignee

• International Business Machines Corporation · US

Inventors

• Barry E. Willner · Ossining, US
• Edith H. Stern · Yorktown Heights, US
• Patrick J. O'Sullivan · Dublin, IE
• Robert C. Weir · Westford, US
• Sean Callanan · Dublin, IE

Key dates

• Filing date: Jul 16, 2007
• Grant date: Jan 13, 2015
• Expiry date: May 29, 2032

Classification

• Technology area (CPC H): Electricity
• CPC primary: H04L63/126
• WIPO field: IT methods for management
• WIPO sector: Electrical engineering

Abstract

A system and method for detecting click fraud where data related to at least one vector associated with a path taken by movement of a cursor on a web page prior to an advertisement on the web page being clicked by a controller of the cursor is received. Click fraud is detected based on an analysis of the received data. A pattern in the at least one vector is analyzed to determine if the cursor movement is human-like and a click fraud detected if the cursor movement is not human-like. In another embodiment, the at least one vector associated with the path is stored and compared with previously stored vectors associated with the path and a click fraud detected if the stored at least one vector is analogous to at least one of the previously stored vectors associated with the path.