Detection of phishing attacks using similarity analysis
A computerized system and method to detect phishing cyber-attacks is described. The approach entails analyzing one or more displayable images of a webpage referenced by a URL to ascertain whether the one or more displayable images, and thus the webpage and potentially an email including the URL, are part of a phishing cyber-attack.
1. A computerized method for detecting a phishing cyber-attack, the method comprising:
inspecting a source code of an object to detect whether the object includes an input prompt in the source code;
extracting a plurality of images from the object if the source code of the object includes the input prompt or discontinuing analysis of the object if the source code is devoid of the input prompt;
generating one or more properties associated with each of the plurality of images included as part of the object;
correlating the one or more properties associated with each image of the plurality of images with one or more properties associated with images of known phishing objects to classify the object; and
generating and issuing an alert to indicate a classification of the object as part of a phishing attack.
2. The computerized method of claim 1 , wherein the inspecting of the source code includes conducting a static analysis of the object.
3. The computerized method of claim 1 , wherein the input prompt includes a displayable element that prompts a user to enter data.
4. The computerized method of claim 1 , wherein the object corresponds to content of a webpage that includes source code associated with the webpage.
5. The computerized method of claim 4 , wherein each of the plurality of images corresponds to a graphical representation of a subset of the webpage.
6. The computerized method of claim 5 , wherein the graphical representation includes one or more embedded images, the one or more embedded images include a logo.
7. The computerized method of claim 1 , wherein the correlating of the one or more properties associated with each image of the plurality of images with the one or more properties associated with the images of known phishing objects comprises assigning a score based on a level of correlation between (i) each property of the one or more properties associated with the object and (ii) properties associated with images of the known phishing objects.
8. A computerized method for detecting a phishing cyber-attack, the method comprising:
extracting a plurality of images from an object;
generating one or more properties associated with each of the plurality of images included as part of the object;
correlating the one or more properties associated with each image of the plurality of images with one or more properties associated with images of known phishing objects to classify the object; and
generating and issuing an alert to indicate a classification of the object as part of a phishing attack,
wherein the correlating of the one or more properties associated with each image of the plurality of images with the one or more properties associated with the images of known phishing objects comprises assigning a score for an image of the plurality of images extracted from the object based on a combination of scores for each of the one or more properties associated with the image.
9. The computerized method of claim 8 , wherein the correlating of the one or more properties associated with each image of the plurality of images with the one or more properties associated with the images of known phishing objects further comprises assigning a score for the object based on a combination of scores assigned to the plurality of images extracted from the object.
10. The computerized method of claim 1 , wherein the one or more properties associated with an image of the plurality of images comprise a cryptographic hash of the image.
11. The computerized method of claim 1 , wherein the one or more properties associated with an image of the plurality of images comprise a perceptual hash of the image, the perceptual hash operates as a fingerprint of the image and is derived from features of the image.
12. The computerized method of claim 11 , wherein the perceptual hash is used in determining a similarity between the image and at least one image of the known phishing objects.
13. A non-transitory persistent storage including logic that, when executed by one or more processors, is configured to detect phishing cyber-attacks, comprising:
logic that, when in operation, conducts a static analysis of the object to inspect a source code of the object to detect whether the source code includes an input prompt and extracts a plurality of images from an object if the source code of the object includes the input prompt or discontinues subsequent analysis of the object if the source code of the object is devoid of the input prompt;
feature analyzer logic that, when in operation, generates one or more properties associated with each of the plurality of images included as part of the object;
classifier logic that, when in operation, correlates the one or more properties associated with each image of the plurality of images with one or more properties associated with images of known phishing objects to classify the object; and
a reporting engine that, when in operation, generates and issues an alert to indicate a classification of the object as part of a phishing attack.
14. The non-transitory persistent storage of claim 13 , wherein the input prompt includes a displayable element that prompts a user to enter data.
15. The non-transitory persistent storage of claim 13 , wherein the object corresponds to content of a webpage that includes source code associated with the webpage.
16. The non-transitory persistent storage of claim 15 , wherein each of the plurality of images corresponds to a graphical representation of a subset of the webpage.
17. The non-transitory persistent storage of claim 16 , wherein the graphical representation includes one or more embedded images, the one or more embedded images include a logo.
18. The non-transitory persistent storage of claim 13 , wherein the classifier logic is configured to attempt to correlate the one or more properties associated with each image of the plurality of images with the one or more properties associated with the images of known phishing objects by at least assigning a score based on a level of correlation between (i) each property of the one or more properties associated with the object and (ii) properties associated with images of the known phishing objects.
19. A non-transitory persistent storage including logic that, when executed by one or more processors, is configured to detect phishing cyber-attacks, comprising:
logic that, when in operation, extracts a plurality of images from an object;
feature analyzer logic that, when in operation, generates one or more properties associated with each of the plurality of images included as part of the object;
classifier logic that, when in operation, correlates the one or more properties associated with each image of the plurality of images with one or more properties associated with images of known phishing objects to classify the object; and
a reporting engine that, when in operation, generates and issues an alert to indicate a classification of the object as part of a phishing attack,
wherein the classifier logic is configured to attempt to correlate the one or more properties associated with each image of the plurality of images with the one or more properties associated with the images of known phishing objects by at least assigning a score for an image of the plurality of images extracted from the object based on a combination of scores for each of the one or more properties associated with the image.
20. The non-transitory persistent storage of claim 19 , wherein the classifier logic is configured to attempt to correlate the one or more properties associated with each image of the plurality of images with the one or more properties associated with the images of known phishing objects by at least assigning the score for the object based on a combination of scores assigned to the plurality of images extracted from the object in lieu of each of the one or more properties associated with the image.
21. The non-transitory persistent storage of claim 13 , wherein the one or more properties associated with an image of the plurality of images comprise a cryptographic hash of the image.
22. The non-transitory persistent storage of claim 19 , wherein the one or more properties associated with an image of the plurality of images comprise a perceptual hash of the image, the perceptual hash operates as a fingerprint of the image and is derived from features of the image.
23. The non-transitory persistent storage of claim 22 , wherein the perceptual hash is used in determining a similarity between the image and at least one image of the known phishing objects.