Title
RiskWriter - Predicting Cyber Risk of an Enterprise.
Abstract
Empirically measuring security posture of an enterprise is a challenging problem. One has to thoroughly understand external and, internal exposure for a given firm to assess security posture at a given time. Various security metrics are used to model each type of security exposure. Due to the lack of data on internal security metrics for a broad sample of firms, the research community has relied on external, proxy data points to assess the cyber risk of a firm. Recent studies, which attempted to solve this problem either used a small set of enterprises or used artificial datasets. Moreover, we are not aware of any existing approach to assess the security posture of an enterprise using only external and business data. In this paper, we present RiskWriter, a framework to assess the internal security posture of an enterprise using only external and business data. In our study, we measure a set of internal, external and business attributes of around 200,000 firms of different sizes, line of business, locations and security profiles for a period of 12 months. Prediction models were built by deriving, for each company, a comprehensive set of metrics using novel filtering and, normalizing techniques and then building machine learning models to assess the internal security posture of a company using only external and business data. We also evaluate RiskWriter with 2000 enterprises, with a variety of metrics and show that prediction is stable with high accuracy. Specifically for this work, the longitudinal study a broad sample of firms and for a period of one year is done for the first time.
Year
DOI
Venue
2018
10.1007/978-3-030-05171-6_5
ICISS
Field
DocType
Citations 
Data science,Proxy (climate),Data mining,Computer science,Filter (signal processing),Predictive modelling,Small set,Internal security,Line of business,Business data
Conference
1
PageRank 
References 
Authors
0.36
6
3
Name
Order
Citations
PageRank
K. Aditya111.03
Slawomir Grzonkowski2718.03
Nhien-An Le-Khac322449.63