Determining a risk probability of a URL using machine learning of URL segments
Systems, methods, and apparatuses enable a machine learning model to determine a risk probability of a URL. A query configurator receives a URL in a query and normalizes the URL. The normalized URL is segmented into a plurality of segments. The plurality of segments is serially provided to the machine learning model trained to provide an indication of risk associated with the URL. The indication of risk associated with the URL can be a probability value based on one or more risk probabilities determined for segment-segment transitions of the URL. A security service compares the probability value of the URL to a threshold value and performs a security action based on a result of comparing the probability value to the threshold value.
1. A computer-implemented method comprising:
receiving a URL in a query;
normalizing the URL into a normalized URL;
segmenting the normalized URL into a plurality of segments;
serially providing each of the plurality of segments into a machine learning model trained to provide an indication of risk associated with the URL;
receiving at least one probability value from the machine learning model;
comparing the probability value to a threshold value; and
performing a security action based on a result of comparing the probability value to the threshold value.
2. The computer-implemented method of claim 1 , wherein serially providing each of the plurality of segments into the machine learning model includes the machine learning model performing time-series predictions of the plurality of segments of the normalized URL.
3. The computer-implemented method of claim 1 , wherein the threshold value is determined based on a security policy.
4. The computer-implemented method of claim 1 , further comprising:
modifying the threshold value in response to probability values of one or more URLs being above the threshold value.
5. The computer-implemented method of claim 1 , wherein the probability value being above the threshold value is indicative of the URL being randomly-generated.
6. The computer-implemented method of claim 1 , wherein serially providing each of the plurality of segments into the machine learning model includes sending each of the plurality of segments to the machine learning model.
7. One or more non-transitory computer-readable storage media storing instructions which, when executed by one or more hardware processors, cause performance of a method comprising:
receiving a URL in a query;
normalizing the URL into a normalized URL;
segmenting the normalized URL into a plurality of segments;
serially providing each of the plurality of segments into a machine learning model trained to provide an indication of risk associated with the URL;
receiving at least one probability value from the machine learning model;
comparing the probability value to a threshold value; and
performing a security action based on a result of comparing the probability value to the threshold value.
8. The one or more non-transitory computer-readable storage media of claim 7 , wherein serially providing each of the plurality of segments into the machine learning model includes the machine learning model performing time-series predictions of the plurality of segments of the normalized URL.
9. The one or more non-transitory computer-readable storage media of claim 7 , wherein the threshold value is determined based on a security policy.
10. The one or more non-transitory computer-readable storage media of claim 7 , further comprising:
modifying the threshold value in response to probability values of one or more URLs being above the threshold value.
11. The one or more non-transitory computer-readable storage media of claim 7 , wherein the probability value being above the threshold value is indicative of the URL being randomly-generated.
12. The one or more non-transitory computer-readable storage media of claim 7 , wherein serially providing each of the plurality of segments into the machine learning model includes sending each of the plurality of segments to the machine learning model.
13. An apparatus comprising:
one or more hardware processors;
memory coupled to the one or more hardware processors, the memory storing instructions which, when executed by the one or more hardware processors, causes the apparatus to:
receive a URL in a query;
normalize the URL into a normalized URL;
segment the normalized URL into a plurality of segments;
serially provide each of the plurality of segments into a machine learning model trained to provide an indication of risk associated with the URL;
receive at least one probability value from the machine learning model;
compare the probability value to a threshold value; and
perform a security action based on a result of comparing the probability value to the threshold value.
14. A computer-implemented method comprising:
receiving a URL from a training database containing a plurality of URLs to train a machine learning model;
normalizing the URL into a normalized URL;
segmenting the normalized URL into a plurality of segments; and
serially providing each of the plurality of segments into the machine learning model, the machine learning model determining an indication of risk associated with each of the plurality of segments.
15. The computer-implemented method of claim 14 , wherein normalizing the URL into the normalized URL comprises:
applying character restrictions to the URL to remove restricted characters.
16. The computer-implemented method of claim 14 , wherein normalizing the URL into the normalized URL comprises:
rearranging portions of the URL.
17. One or more non-transitory computer-readable storage media storing instructions which, when executed by one or more hardware processors, cause performance of a method comprising:
receiving a URL from a training database containing a plurality of URLs to train a machine learning model;
normalizing the URL into a normalized URL;
segmenting the normalized URL into a plurality of segments; and
serially providing each of the plurality of segments into the machine learning model, the machine learning model determining an indication of risk associated with each of the plurality of segments.
18. The one or more non-transitory computer-readable storage media of claim 17 , wherein normalizing the URL into the normalized URL comprises:
applying character restrictions to the URL to remove restricted characters.
19. The one or more non-transitory computer-readable storage media of claim 17 , wherein normalizing the URL into the normalized URL comprises:
rearranging portions of the URL.
20. An apparatus comprising:
one or more hardware processors;
memory coupled to the one or more hardware processors, the memory storing instructions which, when executed by the one or more hardware processors, causes the apparatus to:
receive a URL from a training database containing a plurality of URLs to train a machine learning model;
normalize the URL into a normalized URL;
segment the normalized URL into a plurality of segments; and
serially provide each of the plurality of segments into the machine learning model, the machine learning model determining an indication of risk associated with each of the plurality of segments.