Machine Learning–based Cyber Attacks Targeting on Controlled Information

Miao, Yuantian; Chen, Chao; Pan, Lei; Han, Qing‐Long; Zhang, Jun; Xiang, Yang

doi:10.1145/3465171

Cited by 75 publications

(24 citation statements)

References 82 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Then, inspired by Wang et al, 18 in GCN 1 , that is, the first GCN layer, we generate an aggregation feature

a_{i}

for each f i by utilizing the following aggregation function (

1 \leq c \leq q_{i}

a_{i} script= scriptH (F_{i}) = \sum_{f_{c}^{i} \in F_{i}} (\frac{ReLU ({boldW}_{1} {\overset{f}{true}}_{c}^{i} + {boldb}_{1})}{J ({\overset{f}{true}}_{c}^{i}, f_{i})}),

where

{\tilde{boldf}}_{c}^{i}

is the initial feature of

f_{c}^{i}

, rectified linear unit (ReLU) 37 is a nonlinear activation function, and

W_{1}

and

b_{1}

are the two trained parameters.

J ()(, {\tilde{boldf}}_{c}^{i}, f_{i})

is the graph Laplacian norm 18,38 :

J ({\tilde{boldf}}_{c}^{i}, f_{i}) = {(d (f_{c}^{i}) \cdot d (f_{i}))}^{- 1 / 2},

where

scriptd script(f_{i})

is the degree of node

f_{i}

. We consider

a_{i}

and

{\tilde{boldf}}_{i}

(the initial feature of

f_{i}

) as an input and use an FC layer (FC 1 ) to generate

f_{i}

's derivation feature

d_{i}

: …”

Section: Proposed Tkdmmentioning

confidence: 99%

See 1 more Smart Citation

A two‐phase knowledge distillation model for graph convolutional network‐based recommendation

Huang

Lin

Gong

et al. 2022

Int J of Intelligent Sys

View full text Add to dashboard Cite

Graph convolutional network (GCN)‐based recommendation has recently attracted significant attention in the recommender system community. Although current studies propose various GCNs to improve recommendation performance, existing methods suffer from two main limitations. First, user–item interaction data is generally sparse in practice, highlighting these methods' ineffectiveness in learning user and item feature representations. Second, they usually perform a dot‐product operation to model and calculate user preferences on items, leading to inaccurate user preference learning. To address these limitations, this study adopts a design idea that sharply differs from existing works. Specifically, we introduce the knowledge distillation concept into GCN‐based recommendation and propose a two‐phase knowledge distillation model (TKDM) improving recommendation performance. In Phase I, a self‐distillation method on a graph auto‐encoder learns the user and item feature representations. This auto‐encoder employs a simple two‐layer GCN as an encoder and a fully connected layer as a decoder. On this basis, in Phase II, a mutual‐distillation method on a fully connected layer is introduced to learn user preferences on items with triple‐based Bayesian personalized ranking. Extensive experiments on three real‐world data sets demonstrate that TKDM outperforms classic and state‐of‐the‐art methods related to GCN‐based recommendation problems.

show abstract

“…Then, inspired by Wang et al, 18 in GCN 1 , that is, the first GCN layer, we generate an aggregation feature

a_{i}

for each f i by utilizing the following aggregation function (

1 \leq c \leq q_{i}

a_{i} script= scriptH (F_{i}) = \sum_{f_{c}^{i} \in F_{i}} (\frac{ReLU ({boldW}_{1} {\overset{f}{true}}_{c}^{i} + {boldb}_{1})}{J ({\overset{f}{true}}_{c}^{i}, f_{i})}),

where

{\tilde{boldf}}_{c}^{i}

is the initial feature of

f_{c}^{i}

, rectified linear unit (ReLU) 37 is a nonlinear activation function, and

W_{1}

and

b_{1}

are the two trained parameters.

J ()(, {\tilde{boldf}}_{c}^{i}, f_{i})

is the graph Laplacian norm 18,38 :

J ({\tilde{boldf}}_{c}^{i}, f_{i}) = {(d (f_{c}^{i}) \cdot d (f_{i}))}^{- 1 / 2},

where

scriptd script(f_{i})

is the degree of node

f_{i}

. We consider

a_{i}

and

{\tilde{boldf}}_{i}

(the initial feature of

f_{i}

) as an input and use an FC layer (FC 1 ) to generate

f_{i}

's derivation feature

d_{i}

: …”

Section: Proposed Tkdmmentioning

confidence: 99%

“…(1) where f ̃c i is the initial feature of f c i , rectified linear unit (ReLU) 37 is a nonlinear activation function, and W 1 and b 1 are the two trained parameters. J f f ( ̃, ) c i i is the graph Laplacian norm 18,38 :…”

Section: Feature Representation Learningmentioning

confidence: 99%

A two‐phase knowledge distillation model for graph convolutional network‐based recommendation

Huang

Lin

Gong

et al. 2022

Int J of Intelligent Sys

View full text Add to dashboard Cite

show abstract

“…The popularization of network use also increases the risk of a network attack, bringing many security risks. 1,2 Network security situation assessment (NSSA) can build an appropriate model according to related security incidents, and then assess the threat degree of the entire network system. 3 Since Bass 4 proposed the concept of NSSA, it has been the main topic in situation awareness research.…”

Section: Introductionmentioning

confidence: 99%

Network security situation assessment with network attack behavior classification

Yang

Zhang

Xie

et al. 2022

Int J of Intelligent Sys

View full text Add to dashboard Cite

To solve the problems that existing network security situation assessment (NSSA) methods are difficult to extract features and have poor timeliness, an NSSA method with network attack behavior classification (NABC) is proposed. First, an NABC model is designed. The model combines features and advantages of a parallel feature extraction network (PFEN), a bidirectional gate recurrent unit (BiGRU), and the attention mechanism (ATT). The PFEN module is composed of parallel sparse autoencoders which extract key data from different network attack behaviors. The BiGRU module gets the time-series relationship from the state of three different time periods, finds potential representation rules from network attack behaviors. The ATT module pays more attention to the network traffic key information and improves the NABC accuracy. Second, the NABC detects and classifies attacks from network behaviors, the occurrence number of each attack behavior, and the error probability matrix are counted. Finally, the occurrence number of each attack behavior is corrected according to the error probability matrix, and the network security situation value is calculated through combining the severity factor of each attack behavior. The experimental results show that the precision and recall of the NABC model are improved by 5.28% and 5.65%, respectively, compared with the conventional method. The comparison experiment with the

show abstract

“…For example, in 2020, multiple popular Android apps, such as Baidu Maps, were found leaking users' location data. Thus, data leakage has become one of the most representative security incidents in the past few years [160] [121] [85] [35] [90] [135] [46] [95]. And it poses a significant threat on user's identity security, financial security, or even life security.…”

Section: Introductionmentioning

confidence: 99%

Information flow based defensive chain for data leakage detection and prevention: a survey

Xi,

Chen,

Zhang

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Mobile and IoT applications have greatly enriched our daily life by providing convenient and intelligent services. However, these smart applications have been a prime target of adversaries for stealing sensitive data. It poses a crucial threat to users' identity security, financial security, or even life security. Research communities and industries have proposed many Information Flow Control (IFC) techniques for data leakage detection and prevention, including secure modeling, type system, static analysis, dynamic analysis, etc. According to the application's development life cycle, although most attacks are conducted during the application's execution phase, data leakage vulnerabilities have been introduced since the design phase. With a focus on lifecycle protection, this survey reviews the recent representative works adopted in different phases. We propose an information flow based defensive chain, which provides a new framework to systematically understand various IFC techniques for data leakage detection and prevention in Mobile and IoT applications. In line with the phases of the application life cycle, each reviewed work is comprehensively studied in terms of technique, performance, and limitation. Research challenges and future directions are also pointed out by consideration of the integrity of the defensive chain. CCS Concepts: • Security and privacy → Software security engineering; Logic and verification; Information flow control;• Software and its engineering → Software development process management.

show abstract

Machine Learning–based Cyber Attacks Targeting on Controlled Information

Cited by 75 publications

References 82 publications

A two‐phase knowledge distillation model for graph convolutional network‐based recommendation

A two‐phase knowledge distillation model for graph convolutional network‐based recommendation

Network security situation assessment with network attack behavior classification

Information flow based defensive chain for data leakage detection and prevention: a survey

Contact Info

Product

Resources

About