SenseInput: An Image-Based Sensitive Input Detection Scheme for Phishing Website Detection

Lin, Shih‐Chun; Wl, Pang-Cheng; Chen, Hong-Yen; Morikawa, Tomohiro; Takahashi, Takeshi; Lin, Tsung-Nan

doi:10.1109/icc45855.2022.9838653

Cited by 6 publications

(2 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In article spinning which aims to create the deceitful appearance of what appears to be new content so as to distinguish it from what already exists, Shahid et al [95] utilized some basic lexical features (i.e., word count and sentence count) of generated articles to detect spun content and its seed without needing the text spinner's dictionary. The length of an article is also used to identify a malicious website for phishing and spam [74], [84], [90].…”

Section: A a Taxonomy Of Ctimentioning

confidence: 99%

Understanding and Analyzing Appraisal Systems in the Underground Marketplaces

Li,

Liao

2024

Proceedings 2024 Network and Distributed System Security Symposium

View full text Add to dashboard Cite

An appraisal system is a feedback mechanism that has gained popularity in underground marketplaces. This system allows appraisers, who receive free samples from vendors, to provide assessments (i.e., appraisal reviews) for products in underground marketplaces. In this paper, we present the first measurement study on the appraisal system within underground marketplaces. Specifically, from 17M communication traces from eight marketplaces spanning from Feb 2006 to Mar 2023, we discover 56,229 appraisal reviews posted by 18,701 unique appraisers. We look into the appraisal review ecosystem, revealing five commonly used requirements and merits in the appraiser selection process. These findings indicate that the appraisal system is a well-established and structured process within the underground marketplace ecosystem. Furthermore, we reveal the presence of high-quality and unique cyber threat intelligence (CTI) in appraisal reviews. For example, we identify the geolocations of followers for a social booster and programming languages used for malware. Leveraging our extraction model, which integrates 41 distinct types of CTI, we capture 23,978 artifacts associated with 16,668 (50.2%) appraisal reviews. In contrast, artifacts are found in only 8.9% of listings and 2.7% of non-appraisal reviews. Our study provides valuable insights into this under-explored source of CTI, complementing existing research on threat intelligence gathering.

show abstract

Section: A a Taxonomy Of Ctimentioning

confidence: 99%

Understanding and Analyzing Appraisal Systems in the Underground Marketplaces

Li,

Liao

2024

Proceedings 2024 Network and Distributed System Security Symposium

View full text Add to dashboard Cite

show abstract

“…Their GAN consists of a CNN discriminator and an LSTM generator. Lin et al [13] developed a type of method in which the presence of sensitive inputs in web pages is checked along with the character-level representation of URLs to determine whether they are genuine or phishing. They used the faster-RCNN model to detect sensitive inputs from screenshots of webpages using object detection.…”

Section: Introductionmentioning

confidence: 99%

Deep learning based phishing website detection

Subhashini,

Banerjee,

Kumar

et al. 2023

TELKOMNIKA

View full text Add to dashboard Cite

Phishing attacks use fraudulent websites that trick people into disclosing sensitive information. More effective and precise methods are required to identify phishing websites so that people and organisations can be protected from the damaging effects of these online threats. The aim of this work is to develop a model that can identify phishing uniform resource locator (URLs) more accurately than current approaches while requiring less training time, testing time, and storage space. This research work proposes a novel method for identifying phishing websites using a long short-term memory (LSTM) gated recurrent unit (GRU) algorithm to detect phishing URLs. The accuracy of the suggested method is 98.89%, which is significantly better than the findings of earlier studies. The model also showed a need for shorter training and testing time, and a reduced amount of storage space.

show abstract

A cyber defense system against phishing attacks with deep learning game theory and LSTM-CNN with African vulture optimization algorithm (AVOA)

Elberri,

Tokeşer,

Rahebi

et al. 2024

Int. J. Inf. Secur.

View full text Add to dashboard Cite

Phishing attacks pose a significant threat to online security, utilizing fake websites to steal sensitive user information. Deep learning techniques, particularly convolutional neural networks (CNNs), have emerged as promising tools for detecting phishing attacks. However, traditional CNN-based image classification methods face limitations in effectively identifying fake pages. To address this challenge, we propose an image-based coding approach for detecting phishing attacks using a CNN-LSTM hybrid model. This approach combines SMOTE, an enhanced GAN based on the Autoencoder network, and swarm intelligence algorithms to balance the dataset, select informative features, and generate grayscale images. Experiments on three benchmark datasets demonstrate that the proposed method achieves superior accuracy, precision, and sensitivity compared to other techniques, effectively identifying phishing attacks and enhancing online security.

show abstract

SenseInput: An Image-Based Sensitive Input Detection Scheme for Phishing Website Detection

Cited by 6 publications

References 16 publications

Understanding and Analyzing Appraisal Systems in the Underground Marketplaces

Understanding and Analyzing Appraisal Systems in the Underground Marketplaces

Deep learning based phishing website detection

A cyber defense system against phishing attacks with deep learning game theory and LSTM-CNN with African vulture optimization algorithm (AVOA)

Contact Info

Product

Resources

About