CySecBERT: A Domain-Adapted Language Model for the Cybersecurity Domain

Bayer, М.; Küehn, Philipp; Shanehsaz, Ramin; Reuter, Christian

doi:10.48550/arxiv.2212.02974

Cited by 2 publications

(2 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These approaches have been recently also applied for the definition of CS NLP-based methodologies, allowing for the automatic analysis of NL documents. The authors of [38] presented a BERT model devoted to provide CS domain embeddings for analysing CS texts. They pretrained the model on a large corpus consisting of the text extracted from scientific papers, Twitter posts, CS-domain web pages, and vulnerability database, successfully testing it on several CS NLP tasks.…”

Section: Nlp For Unstructured Cyber Security Text Analysismentioning

confidence: 99%

Cyber threat assessment and management for securing healthcare ecosystems using natural language processing

Silvestri,

Islam,

Amelin

et al. 2023

Int. J. Inf. Secur.

View full text Add to dashboard Cite

The healthcare sectors have constantly faced significant challenge due to the rapid rise of cyber threats. These threats can pose any potential risk within the system context and disrupt the critical healthcare service delivery. It is therefore necessary for the healthcare organisations to understand and tackle the threats to ensure overall security and resilience. However, threats are continuously evolved and there is large amount of unstructured security-related textual information is available. This makes the threat assessment and management task very challenging. There are a number of existing works that consider Machine Learning models for detection and prediction of cyber attack but they lack of focus on the Natural Language Processing (NLP) to extract the threat information from unstructured security-related text. To this end, this work proposes a novel method to assess and manage threats by adopting natural language processing. The proposed method has been tailored for the healthcare ecosystem and allows to identify and assess the possible threats within healthcare information infrastructure so that appropriate control and mitigation actions can be taken into consideration to tackle the threat. In detail, NLP techniques are used to extract the useful threat information related to specific assets of the healthcare ecosystems from the largely available security-related information on Internet (e.g. cyber security news), to evaluate the level of the identified threats and to select the required mitigation actions. We have performed experiments on real healthcare ecosystems in Fraunhofer Institute for Biomedical Engineering, considering in particular three different healthcare scenarios, namely implantable medical devices, wearables, and biobank, with the purpose of demonstrating the feasibility of our approach, which is able to provide a realistic manner to identify and assess the threats, evaluate the threat level and suggest the required mitigation actions.

show abstract

Section: Nlp For Unstructured Cyber Security Text Analysismentioning

confidence: 99%

Cyber threat assessment and management for securing healthcare ecosystems using natural language processing

Silvestri,

Islam,

Amelin

et al. 2023

Int. J. Inf. Secur.

View full text Add to dashboard Cite

show abstract

“…These approaches have been recently also applied for the definition CS NLP-based methodologies, allowing for the automatic analysis of NL documents. The authors of [34] presented a BERT model devoted to provide CS domain embeddings for analysing CS texts. They pretrained the model on a large corpus consisting of the text extracted from scientific papers, Twitter posts, CS-domain webpages, and vulnerability database, successfully testing it on several CS NLP tasks.…”

Section: Nlp For Unstructured Cyber Security Text Analysismentioning

confidence: 99%

Cyber Threat Assessment and Management for Securing Healthcare Ecosystems using Natural Language Processing

Silvestri,

Islam,

Amelin

et al. 2023

Preprint

View full text Add to dashboard Cite

The healthcare sectors have constantly faced significant challenge due to the rapid rise of cyber threats. These threats can pose any potential risk within the system context and disrupt the critical healthcare service delivery. It is therefore necessary for the healthcare organisations to understand and tackle the threats to ensure overall security and resilience. However, threats are continuously evolved and there is large amount of unstructured security related textual information is available. This makes the threat assessment and management task very challenging. There are a number of existing works that consider Machine Learning models for detection and prediction of cyber attack but they lack of focus on the Natural Language Processing (NLP) to extract the threat information from unstructured security related text. To this end, this work proposes a novel method to assess and manage threats by adopting natural language processing. The proposed method has been tailored for the healthcare ecosystem and allows to identify and assess the possible threats within healthcare information infrastructure so that appropriate control and mitigation actions can be taken into consideration to tackle the threat. In detail, NLP techniques are used to extract the useful threat information related to specific assets of the healthcare ecosystems from the largely available security related information on Internet (e.g., cyber security news), to evaluate the level of the identified threats and to select the required mitigation actions. We have performed experiments on real healthcare ecosystems in Fraunhofer Institute for Biomedical Engineering (IBMT), considering in particular three different healthcare scenarios, namely implantable medical devices, wearables, and biobank, with the purpose of demonstrating the feasibility of our approach, which is able to provide a realistic manner to identify and assess the threats, evaluate the threat level and suggest the required mitigation actions.

show abstract

CySecBERT: A Domain-Adapted Language Model for the Cybersecurity Domain

Cited by 2 publications

References 0 publications

Cyber threat assessment and management for securing healthcare ecosystems using natural language processing

Cyber threat assessment and management for securing healthcare ecosystems using natural language processing

Cyber Threat Assessment and Management for Securing Healthcare Ecosystems using Natural Language Processing

Contact Info

Product

Resources

About