Exploit Prediction using Twitter Analytics

Select a vulnerability from the dropdown menu:

Details for this vulnerability:



Timeseries plot for this vulnerability:


The list contains vulnerabilities from 2014 and 2015 that are known to be exploited in the wild.

After a vulnerability is selected from the dropdown menu, the page displays basic information and a time series for the selected entry.

The basic information includes the CVE ID (a unique identifier for that vulnerability) and a brief description of the vulnerability.

The plot shows a time series. The series are: the total tweeting volume, the retweeting volume for a subset of informative users, and the confidence of the Twitter-based exploit detector in marking the vulnerability as exploited.

The plot also marks the point in time where the Symantec Antivirus solutions are updated with the associated virus definitions.

The left vertical axis corresponds to the number of tweets, while the right vertical axis corresponds to the classifier confidence percentage.

You can view important tweets that pop out at the corresponding posting date by moving the cursor over the plot.

You can zoom in or out using the mouse scroll while the pointer is over the plot region. While zoomed into a region, you can drag the plot left or right to change the timeframe.

For viewing a particular series, move the cursor over the series description in the legend located at the bottom of the plot.


This website presents results from our paper:

Vulnerability Disclosure in the Age of Social Media: Exploiting Twitter for Predicting Real-World Exploits

For any questions, please contact us:

Tudor Dumitras - tdumitra at umiacs dot umd dot edu

Octavian Suciu - osuciu at umiacs dot umd dot edu