Back to results
US 9,544,272 B2Granted
Detecting image spam
AssigneeIntel Corporation
InventorsDmitri Alperovitch, Nick Black, Jeremy Gould, Paul Judge, Sven Krasser, Phyllis Adele Schneck, Yuchun Tang, Aarjav Jyotindra Neeta Trivedi, Lamar Lorenzo Willis, Weilai Yang, Jonathan Alexander Zdziarski
Life of this patent
Estimating the term…
Abstract
Methods and systems for operation upon one or more data processors for detecting image spam by detecting an image and analyzing the content of the image to determine whether the incoming communication comprises an unwanted communication.
Claims20
At least one non-transitory machine accessible storage medium having code stored thereon, the code when executed on a machine, cause the machine to:
- access received data packets associated with a particular communication from a particular source on a network;
- parse the data packets to identify image data included in the particular communication;
- determine similarities between the image data of the particular communication and content of a plurality of other communications from a plurality of other sources, wherein the determining similarities comprises determining whether images similar to the image data are included in content of one or more of the other communications, and some of the plurality of other sources have reputable source reputation scores and some of the plurality of other sources have non-reputable source reputation scores;
- determine a source reputation score for the particular source, wherein the source reputation score comprises a respective category score of the particular source in a plurality of categories, the plurality of categories comprises a spam category, the source reputation score is calculated based at least in part on determined similarities between the image data and the content of the plurality of other communications, and the source reputation score for the particular source is determined based on both the reputable source reputation scores and the non-reputable source reputation scores in the source reputation scores of the plurality of sources;
- and
- cause the particular communication to be processed based on the reputation score for the particular source.
3 more elements
A method comprising:
- receiving data packets associated with a particular communication from a particular source on a network;
- parsing the data packets to identify image data included in content of a particular communication;
- determining similarities between the content of the particular communication and content of a plurality of other communications from a plurality of other sources, wherein determining similarities comprises determining whether images similar to the image data are included in one or more of the other communications, and some of the plurality of other sources have reputable source reputation scores and some of the plurality of other sources have non-reputable source reputation scores;
- determining a source reputation score for the particular source based on the similarities and the source reputation scores of the plurality of sources, wherein the source reputation score for the particular source comprises a respective category score of the particular source in a plurality of categories, and the plurality of categories comprises a spam category, and the source reputation score for the particular source is determined based on both the reputable source reputation scores and the non-reputable source reputation scores in the source reputation scores of the plurality of other sources;
- and
- causing the particular communication to be processed based on the reputation score for the particular source.
3 more elements
A system comprising:
- one or more processor devices;
- a storage device;
- components to:
- access received data packets associated with a particular communication from a particular source on a network;
- parse the data packets to identify image data included in a particular communication from a particular source;
- determine similarities between the image data of the particular communication and content of a plurality of other communications from a plurality of other sources, wherein determining similarities comprises determining whether images similar to the image data are included in content of one or more of the other communications, and some of the plurality of other sources have reputable source reputation scores and some of the plurality of other sources have non-reputable source reputation scores;
- identify a source reputation score for the particular source, wherein the source reputation score comprises a respective category score of the particular source in a plurality of categories, the plurality of categories comprises a spam category, the source reputation score is calculated based at least in part on determined similarities between the image data of the particular communication and the content of the plurality of other communications, and the source reputation score for the particular source is determined based on both the reputable source reputation scores and the non-reputable source reputation scores in the source reputation scores of the plurality of sources;
- and
- cause the particular communication to be processed based on the reputation score for the particular source.
6 more elements
Dependencies are inferred from the claim text.