Network Security Research Label Availability Study, 2009-2013

Sebastian Abt from Darmstadt University of Applied Sciences conducted a systematic study of network security research accepted at top IT security conferences between 2009 and 2013. The analysis reveals that 70% of the reviewed papers rely on manually compiled datasets, and only 10% release their data after compilation. This work defines and discusses the characteristics of the missing labelled data problem in the field.

Use Cases

• Analyzing trends in data usage and sharing practices in network security research based on the systematic review of conference papers.
• Studying the prevalence of manually compiled versus public datasets in security research based on the reported 70% figure.
• Investigating barriers to data sharing and ground-truth establishment in the field based on the discussion of the problem.

Strengths

• The study is based on a systematic review of papers from top IT security conferences, providing a focused analysis.
• It provides specific, quantified findings: 70% of papers rely on manual data compilation and only 10% release their data.

Limitations

• The dataset itself (the compiled study results) is not described in detail; column-level documentation is absent.
• The temporal coverage is limited to papers from 2009-2013; findings may not reflect current practices.
• Row count and file format of the underlying study data are unknown, limiting suitability assessment.

Provenance

Source

Sebastian Abt, Darmstadt University of Applied Sciences

Collection Method

Systematic study of network security research papers accepted at top IT security conferences.

Time Range

2009-2013