Building and Evaluating an Annotated Corpus for Automated Recognition of Chat-Based Social Engineering Attacks

Tsinganos, Nikolaos; Mavridis, Ioannis

doi:10.3390/app112210871

Cited by 11 publications

(18 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To achieve her goal, a malicious interlocutor repeats her arguments many times in a conversation to convince and manipulate her partner. This was also confirmed, after processing our CSE corpus [2], where we observed that 83% of social engineering attackers tend to insist on their arguments to exfiltrate the targeted type of critical information. The persistence of an interlocutor to regurgitate the same topic that could lead to sensitive data exfiltration can be considered an enabler of a successful CSE attack.…”

Section: Introductionsupporting

confidence: 80%

“…This approach results in our proposed model, called CSE-PersistenceBERT, which learns a universal representation that transfers knowledge with only minimal adaptation to the paraphrase recognition task. Initially, the CSE-PersistenceBERT model has access to a large corpus of unlabeled text on which BERT has been pre-trained, and later it uses the CSE-Persistence corpus, which emerged from the CSE corpus of our previous work [2]. CSE-Persistence corpus generated after manually annotating the CSE corpus and utilizing the CSE ontology [2].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Applying BERT for Early-Stage Recognition of Persistence in Chat-Based Social Engineering Attacks

Tsinganos

Fouliras²,

Mavridis³

2022

Applied Sciences

Self Cite

View full text Add to dashboard Cite

Chat-based social engineering (CSE) attacks are attracting increasing attention in the Small-Medium Enterprise (SME) environment, given the ease and potential impact of such an attack. During a CSE attack, malicious users will repeatedly use linguistic tricks to eventually deceive their victims. Thus, to protect SME users, it would be beneficial to have a cyber-defense mechanism able to detect persistent interlocutors who repeatedly bring up critical topics that could lead to sensitive data exposure. We build a natural language processing model, called CSE-PersistenceBERT, for paraphrase detection to recognize persistency as a social engineering attacker’s behavior during a chat-based dialogue. The CSE-PersistenceBERT model consists of a pre-trained BERT model fine-tuned using our handcrafted CSE-Persistence corpus; a corpus appropriately annotated for the specific downstream task of paraphrase recognition. The model identifies the linguistic relationship between the sentences uttered during the dialogue and exposes the malicious intent of the attacker. The results are satisfactory and prove the efficiency of CSE-PersistenceBERT as a recognition mechanism of a social engineer’s persistent behavior during a CSE attack.

show abstract

Section: Introductionsupporting

confidence: 80%

Section: Introductionmentioning

confidence: 99%

Applying BERT for Early-Stage Recognition of Persistence in Chat-Based Social Engineering Attacks

Tsinganos

Fouliras²,

Mavridis³

2022

Applied Sciences

Self Cite

View full text Add to dashboard Cite

show abstract

“…After conducting a quantitative analysis to examine the existing persuasion principles in our CSE corpus [41], we concluded that authority and commitment were the most common persuasion principles used by social engineers. Nevertheless, in this study, all persuasion principles were considered to be equally important.…”

Section: B Cialdini's Persuasion Principlesmentioning

confidence: 99%

“…Thus, identifying a persuasive payload in a sentence means identifying informative local features that may be repeated, regardless of where they are placed in the sentence. Let us consider the following sentence that is part of our CSE Corpus [41]: "I need that information to report back to my boss.". We can easily conclude that some of the words are highly informative of a persuasive payload existence (i.e., the word boss denotes a possible use of the persuasion principle of authority), which holds true regardless of the position of this word in the sentence.…”

Section: The Proposed Cse-pucmentioning

confidence: 99%

Utilizing Convolutional Neural Networks and Word Embeddings for Early-Stage Recognition of Persuasion in Chat-Based Social Engineering Attacks

2022

Self Cite

View full text Add to dashboard Cite

Social engineering is widely recognized as the key to successful cyber-attacks. Chat-based social engineering (CSE) attacks are attracting increasing attention because of recent changes in the digital work environment. Sophisticated CSE attacks target human personality traits, and persuasion is regarded as the catalyst to successful CSE attacks. To date, research in social engineering has mostly focused on phishing attacks, neglecting the importance of chat-based software. This paper describes the design and implementation of a persuasion classifier that utilizes machine learning and natural language processing techniques. For this purpose, a convolutional neural network was trained on a chat-based social engineering corpus (CSE Corpus), specifically annotated for recognizing Cialdini's persuasion principles. The proposed persuasion classifier network, named CSE-PUC, can determine whether a sentence carries a persuasive payload by producing a probability distribution over the sentence classes as a persuasion container. The present study is expected to contribute to our understanding of utilizing existing machine learning models and integrating context-aware information into real-life cyber security threats. The experimental application results reported in this work confirm that the approach taken can recognize persuasion methods and is thus able to protect an interlocutor from being victimized.

show abstract

“…Koyun & Al Janabi (2017) [12] Aldawood & Skinner (2019) [13] Many studies focused on natural language processing (NLP) to detect potential SEAs. [14] proposed a chat-based SEA recognition by evaluation of specific purpose written text on social engineering domain. The study enhanced the understanding of in-context features used by attackers.…”

Section: Internationalmentioning

confidence: 99%

Social Engineering Attacks in E-Government System: Detection and Prevention

Ahmed

2022

IJAEML

View full text Add to dashboard Cite

Purpose: E-Government system emerged as a novel public service provision platform that enables governance in an efficient and transparent manner globally. However, despite the success recorded so far by the increase in the use of information and communication technology (ICT) and E-government for public service provision. Social engineering attack (SEA) is one of the challenging information security attacks that prove to be difficult to tackle. This is because the attackers leverage on peoples’ weakness to exploit the system instead of technical vulnerabilities. Design/Methodology/Approach: This paper uses PESTLE (political, economic, social, technology, legal and environment) analysis to critically evaluate the external factors affecting SEAs in E-government system. Findings/Result: The study identified phishing, Baiting, Pretexting, Quid Pro Quo, Honey Trap, Tail Gating, and Pharming as the major SEA techniques used to exploit E-government systems. Furthermore, the author suggest training and awareness programme as the most effective way to detect as well as prevent SEA in E-government system. Users should be aware of the languages with terms requesting urgent response as well as unusual or unexpected situation in a suspicious messages or attachment as factors to detect SEA. Technical controls using natural language processes (NLP), security policies, multifactor authentication (MFA) as well as secured preservation of confidential information from suspicious users are some of the SEA preventive measures. Originality/Value: A flexible and efficient interaction among citizens, businesses and government organizations is a critical factor for successful E-Government system. SEA is one of major challenges affecting communications in E-government system that requires attention. In conclusion, studies toward technological approach for solution of SEA in E-government is recommended. Paper Type: Conceptual Research.

show abstract

Building and Evaluating an Annotated Corpus for Automated Recognition of Chat-Based Social Engineering Attacks

Cited by 11 publications

References 30 publications

Applying BERT for Early-Stage Recognition of Persistence in Chat-Based Social Engineering Attacks

Applying BERT for Early-Stage Recognition of Persistence in Chat-Based Social Engineering Attacks

Utilizing Convolutional Neural Networks and Word Embeddings for Early-Stage Recognition of Persuasion in Chat-Based Social Engineering Attacks

Social Engineering Attacks in E-Government System: Detection and Prevention

Contact Info

Product

Resources

About