| US 7,552,230 B2 | ||
| Method and apparatus for reducing spam on peer-to-peer networks | ||
| David A. George, Somers, N.Y. (US); Raymond B. Jennings, III, Ossining, N.Y. (US); and Jason D. LaVoie, Mahopac, N.Y. (US) | ||
| Assigned to International Business Machines Corporation, Armonk, N.Y. (US) | ||
| Filed on Jun. 15, 2005, as Appl. No. 11/152,890. | ||
| Prior Publication US 2007/0011347 A1, Jan. 11, 2007 | ||
| Int. Cl. G06F 15/16 (2006.01) | ||
| U.S. Cl. 709—235 [709/206; 709/227; 709/232; 713/154; 713/170; 713/181] | 1 Claim |

| 1. A method for reducing spam in a network, said method comprising:
receiving at a first node in said network a search message from a user of said first node, said search message containing
at least one search keyword relating to data for which the user wishes to search;
adding at said first node one or more decoy words to said search message, wherein said one or more decoy words comprise one
or more arbitrarily selected words that are unrelated to said at least one search keyword;
sending by said first node said search message, including said one or more decoy words, to one or more nodes in said network;
receiving at said first node at least one response message in response to said search message;
determining at said first node whether said at least one response message contains enough of said one or more decoy words
to be categorized as spam, wherein said at least one response message contains enough of said one or more decoy words to be
categorized as spam if at least a predefined threshold percentage of a total number of words or characters in said response
message match said one or more decoy words;
identifying, by said first node to said user, at least one response message that is categorized as spam;
categorizing by said first node at least one sender of said at least one response message that is categorized as spam as a
spammer; and
ignoring by said first node subsequent messages sent by said at least one sender for a predefined period of time.
|