BADGERS 2014 - Building Analysis Datasets and Gathering Experience Returns for Security

The BADGERS workshop is the venue for research on “Big Data” for security.
In contrast to the systems community, security researchers have only recently started collecting and looking at Internet-scale, real-world data (e.g., the EU WOMBAT and the US PREDICT initiatives). Experimental security analysis performed on such data is often hampered by concerns such as confidentiality, privacy, and liability. However attackers have become experts in leveraging the whole Internet to achieve their goals. To understand the modus operandi and the motivations of attackers, both the access to Internet-scale, real-world data and the techniques to mine it for relevant security knowledge are necessary. Hence there is a growing need to widen the scope of data-driven security analysis.
The BADGERS workshop is positioned at the confluence of computer security and general purpose large-scale data processing and aims at bringing together people (e.g., researchers, practitioners, system administrators, security analysts) active in the emerging domain of security-related data collection and analysis for Internet-scale computer systems and networks. By giving visibility to existing solutions, the workshop promotes and encourages the better sharing of data and knowledge. The increasing availability of tools and techniques to process large-scale data (aka Big Data) will benefit computer security.
Topics of Interest
The BADGERS workshop seeks papers in areas related to Big Data for security, including the following:
Scalable data collection from networks, hosts, or applications
Real-time gathering and aggregation of diverse sets of raw data
Summarization of raw data with respect to security goals
Attack-resilient data collection
Characterization of dataset external validity
Scalability of security analysis with data volume
Scalability of security analysis with concurrent-attack volume
Combined historical and real-time security analysis
Evaluation of result accuracy for large datasets
Design of realistic large-scale testbeds
Real-time, incremental anonymization for data sharing
Successful, failed, and novel models of data sharing
Sharing of analysis results and supporting data
Internet-scale sharing of security knowledge Legal issues around data collection and sharing
We especially encourage novel and controversial ideas that are not on the above list. Our goal is to create through BADGERS a venue for fruitful discussion about Big Data for security.