Automatic Labeling for Entity Extraction in Cyber Security

TL;DR

The paper presents an automated labeling method for cyber-security entity extraction that leverages domain-specific structured data, enabling rapid training of high-accuracy models with minimal manual annotation.

Contribution

It introduces a novel automatic labeling technique using structured data and provides a publicly available annotated corpus for cyber-security entities.

Findings

Achieved near-perfect precision, recall, and accuracy.

Training times under 17 seconds on a large corpus.

Enabled rapid model development with minimal manual labeling.

Abstract

Timely analysis of cyber-security information necessitates automated information extraction from unstructured text. While state-of-the-art extraction methods produce extremely accurate results, they require ample training data, which is generally unavailable for specialized applications, such as detecting security related entities; moreover, manual annotation of corpora is very costly and often not a viable solution. In response, we develop a very precise method to automatically label text from several data sources by leveraging related, domain-specific, structured data and provide public access to a corpus annotated with cyber-security entities. Next, we implement a Maximum Entropy Model trained with the average perceptron on a portion of our corpus ($\sim$750,000 words) and achieve near perfect precision, recall, and accuracy, with training times under 17 seconds.