Vulnerability Detection Methods Research Articles

Tensor‐based gated graph neural network for automatic vulnerability detection in source code

AbstractThe rapid expansion of smart devices leads to the increasing demand for vulnerability detection in the cyber security field. Writing secure source codes is crucial to protect applications and software. Recent vulnerability detection methods are mainly using machine learning and deep learning. However, there are still some challenges, how to learn accurate source code semantic embedding at the function level, how to effectively perform vulnerability detection using the learned semantic embedding of source code and how to solve the overfitting problem of learning‐based models. In this paper, we consider codes as various graphs with node features and propose a tensor‐based gated graph neural network called TensorGNN to produce code embedding for function‐level vulnerability detection. First, we propose a high‐dimensional tensor for combining different code graph representations. Second, inspired by the work of tensor technology, we propose the TensorGNN model to produce accurate code representations using the graph tensor. We evaluate our model on 7 C and C++ large open‐source code corpus (e.g. SARD&NVD, Debian, SATE IV, FFmpeg, libpng&LibTiff, Wireshark and Github datasets), which contains 13 types of vulnerabilities. Our TensorGNN model improves on existing state‐of‐the‐art works by 10%–30% on average in terms of vulnerability detection accuracy and F1, while our TensorGNN model needs less training time and model parameters. Specifically, compared with other existing works, our model reduces 25–47 times of the number of parameters and decreases 3–10 times of training time. Results of evaluations show that TensorGNN has better performance while using fewer training parameters and less training time.

Vulnerability detection based on federated learning

Context:Detecting potential vulnerabilities is a key step in defending against network attacks. However, manual detection is time-consuming and requires expertise. Therefore, vulnerability detection must require automated techniques. Objective:Vulnerability detection methods based on deep learning need to rely on sufficient vulnerable code samples. However, the problem of code islands has not been extensively researched. For example, in the case of multi-party vulnerability data, how to securely combine multi-party data to improve vulnerability detection performance. From the perspectives of data augmentation and data security, we propose a vulnerability detection framework based on federated learning (VDBFL). VDBFL is a new model for vulnerability code detection that combines multi-party data. Method:Firstly, VDBFL utilizes the code property graph as a code representation. The code property graph contains various semantic dependencies of the code. Secondly, VDBFL utilizes graph neural networks and convolutional neural networks as the code feature extractor. VDBFL utilizes the jump-structured graph attention network to aggregate node information of important neighbors. Finally, VDBFL utilizes horizontal federated learning to train a local vulnerability detection model for the client. Result:In the real world, VDBFL improves F1-Score by 37.4% compared to the vulnerability detection method Reveal. Among the 5401 vulnerability samples, VDBFL detected 11.8 times more vulnerabilities than Reveal. Conclusion:Under different datasets, VDBFL has shown better performance than advanced vulnerability detection methods in multiple metrics. In addition, the federated learning stage of VDBFL can be expanded on top of the feature extraction stage of any vulnerable detection method.

A Static Detection Method for SQL Injection Vulnerability Based on Program Transformation

Static analysis is popular for detecting SQL injection vulnerabilities. However, due to the lack of accurate modeling of object-oriented database extensions, current methods fail to accurately detect SQL injection vulnerabilities in applications that use object-oriented database extensions. We propose a program transformation-based SQL injection vulnerability detection method to address this issue. This method consists of two stages: program transformation and vulnerability detection. In the first stage, object-oriented database extensions are automatically transformed into semantically equivalent procedural database extensions through the identification of key statements, call relation verification, and program transformation. In the second stage, application programs are automatically scanned using a combination of control flow graph construction and taint analysis techniques to detect SQL injection vulnerabilities. Based on the proposed method, we have implemented the OODBE-SCAN prototype system and performed experimental analysis on eight modern PHP applications. We compare OODBE-SCAN with two related static analysis tools, RIPS and Seay. The results show that OODBE-SCAN can detect more real-world vulnerabilities and has higher accuracy than existing methods.

A vulnerability detection algorithm based on residual graph attention networks for source code imbalance (RGAN)

It is essential to detect potential vulnerabilities in software to ensure its safety. As software systems become more complex, traditional static vulnerability detection methods perform poorly. Currently, deep learning-based vulnerability detection models only extract source code vulnerability features using sequences or graphs. Sequential neural networks ignore structural information in the code, such as control flow diagrams and data flow diagrams. Additionally, graph neural networks cannot accurately extract features due to the lack of effective methods for extracting nodes’ features and aggregating global information. To address the above issue, we propose a vulnerability detection algorithm based on residual graph attention networks for source code imbalance (RGAN). Firstly, a local feature extraction module (PE-BL-A module) is designed. Using the sequence neural network, the module extracts various useful features, including node features in a control flow diagram based on local semantic features. Secondly, we present the Residual Graph Attention Network module (RGAT). To learn and update node features along the control flow direction, the module uses a graph attention network with residual connections. In this module, a mean biaffine attention pooling mechanism is proposed that can extract total graph vulnerability features more effectively. Thirdly, a dynamic cross-entropy loss function is designed. Using this function, it can handle sample imbalances during training. Finally, experiments conducted on several benchmark datasets demonstrate that the proposed model achieves state-of-the-art results.

Enhancing vulnerability detection via AST decomposition and neural sub-tree encoding

The explosive growth of software vulnerabilities poses a serious threat to the system security and has become one of the urgent problems of the day. However, existing vulnerability detection methods are still faced with limitations in reaching the balance between detection accuracy, efficiency and applicability. Following a divide-and-conquer strategy, this paper proposes TrVD (abstract syntax Tree decomposition based Vulnerability Detector) to disclose the indicative semantics implied in the source code fragments for accurate and efficient vulnerability detection. To facilitate the capture of subtle semantic features, TrVD converts the AST of a code fragment into an ordered set of sub-trees of restricted sizes and depths with a novel decomposition algorithm. The semantics of each sub-tree can thus be effectively collected with a carefully designed tree-structured neural network. Finally, a Transformer-style encoder is utilized to aggregate the long-range contextual semantics of all sub-trees into a vulnerability-specific vector to represent the target code fragment. The extensive experiments conducted on five large datasets consisting of diverse real-world and synthetic vulnerable samples demonstrate the performance superiority of TrVD against SOTA approaches in detecting the presence of vulnerabilities and pinpointing the vulnerability types. The ablation studies also confirm the effectiveness of TrVD’s core designs.

BinAIV: Semantic-enhanced vulnerability detection for Linux x86 binaries

Binary code vulnerability detection is an important research direction in the field of network security. The extensive reuse of open-source code has led to the spread of vulnerabilities that originally only affected a small number of targets to other software. Existing vulnerability detection methods are mainly based on binary code similarity analysis, that is, by comparing the similarity of code embedding to detect vulnerabilities. However, existing methods lack semantic understanding of binary code and cannot distinguish between different functions with similar code structures, which reduces the accuracy of vulnerability detection. This paper proposes a binary vulnerability detection method BinAIV based on function semantics. BinAIV is based on a neural network model, which defines and constructs binary function semantics to achieve more accurate similarity analysis. Experimental results show that in terms of binary code similarity analysis performance, BinAIV has a significant improvement compared to traditional methods that only use function embedding. In cross-compiler function search, cross-optimization function search, and cross-obfuscation function search experiments, the average Recall@1 value of BinAIV compared to the best-performing baseline methods increased by 40.1 %, 99.8 %, and 184.0 %. In the real-world vulnerability detection experiment, BinAIV had the highest detection accuracy for all vulnerabilities, with an improvement of 155.1 % and 97.7 % compared to Asm2Vec and SAFE, respectively.

VulGAI: vulnerability detection based on graphs and images

Deep learning models have been widely used in the field of vulnerability detection. Deep learning-based vulnerability detection methods can automatically learn code patterns. Some methods consider processing codes as text sequences to achieve scalable vulnerability detection. They leverage natural language processing models to extract code features. These methods do not consider the code's semantic structure and treat code slices as text. Vulnerability detection methods based on graph structures and graph neural networks are more accurate than text-based methods. However, these methods lack scalability in practice. Both graph generation and graph neural network training are all time-consuming. We propose a vulnerability detection method based on graphs and images (VulGAI). VulGAI choose the more reasonable node centrality to generate the image. It can preserve program details and distinguish node importance from different perspectives. In addition, we design a more efficient CNN model, which reduces computational overhead and improves detection performance (Time and Accuracy). We implement VulGAI and evaluate six methods (VulDePecker, SySeVR, Devign, VulCNN, mVulPreter, and VulGAI) on 40,657 functions. Experimental results show that VulGAI achieves higher Accuracy, TPR, and F1-Score than the others. In addition, we compare VulGAI and VulCNN on 30270 real-world functions. VulGAI outperforms VulCNN by 1.48 times in the number of TP. VulGAI is about 3.9 times faster than VulCNN in detection time.

Static vulnerability detection based on class separation

Software vulnerability detection is a key step to prevent the system from being attacked. However, tens of thousands of codes have brought great challenges to engineers, so we urgently need an automatic and intelligent vulnerability detection method. The existing vulnerability detection model based on deep learning has the problem that it is difficult to separate the features of vulnerable and neutral code. Based on the code data drive, this paper proposes a static vulnerability detection method SDV(Statically Detecting Vulnerability) for C∖C++ programs. SDV is a function-level vulnerability code detection method. This paper uses a code property graph to represent the code and decouples the feature extractor and the classifier. In the graph feature extraction stage, we use Jump Graph Attention Network layers and convolutional pooling layers. Their combination can not only prevent the over-smoothing problem but also separate the sample classes deeply. Finally, on the chrdeb dataset, SDV outperforms state-of-the-art function-level vulnerability detection methods by 52.3%, 15.9%, and 39.6% in Precision, Recall, and F1-Score, respectively. On the real project sard, the number of vulnerabilities detected by SDV is 10.7 times more than Reveal.

SCGformer: Smart contract vulnerability detection based on control flow graph and transformer

AbstractThe security of smart contract has always been one of the significant problems in blockchain. As shown in previous studies, vulnerabilities in smart contracts can lead to unpredictable losses. With the rapid growth of the number of smart contracts, more and more data driven detection technologies based on machine learning have been proposed. However, some state‐of‐the‐art approaches mainly rely on the source code of smart contract. These methods are limited by the openness of the source code and the version of the programming language. To address this problem, we propose a novel vulnerability detection method based on transformer by constructing the control flow graph (CFG) of smart contracts operation codes (opcodes), which shields the difference of various versions of program language. Extensive experiments are conducted to evaluate the effectiveness of the proposed method on the authors' own collected dataset. The experimental results show that the proposed method achieves 94.36% accuracy in vulnerability detection, which performs better than other state‐of‐the‐art methods.

DetAC: Approach to Detect Access Control Vulnerability in Web Application Based on Sitemap Model with Global Information Representation

Access control vulnerabilities that lead to elevated privileges are among the most dangerous vulnerabilities in Web applications. Most of the existing detection methods use dynamic or static analysis techniques alone, which suffer from high manual involvement, low automation, high leakage rate, low page coverage, and other deficiencies. To this end, this paper proposes a novel access control vulnerability detection method (DetAC) based on a sitemap model with global information representation. This method first constructs a static site-wide sitemap model based on the page link addresses in the Web application source code through static analysis techniques. After that, the application is logged in and executed dynamically with different role users. During this process, execution traces and request parameters are collected and converted into annotations to fill the corresponding edges of the static site-wide sitemap model. Then, the sitemap model with global information representation is obtained. This model can represent both the global control flow and data flow of the application. Then DetAC analyzes the role-based and user-based access control policies of the Web application based on the node reachability and annotated data features of the model. And according to the information such as role, user, and access resources, it generates attack vectors to achieve different roles and the same role of different users to access each other’s resources. Finally, access control vulnerabilities are detected based on the equivalence of the results obtained using attack vector access and normal access to the Web application server. DetAC was validated on five real open-source Web applications, and the results showed that DetAC successfully detected up to 12 access control vulnerabilities, which are more than those of the traditional seven tools. The dynamic analysis page coverage rate was significantly improved during the detection process, reaching an average of 91.37%.

Smart Contract Vulnerability Detection Based on Deep Learning and Multimodal Decision Fusion.

With the rapid development and widespread application of blockchain technology in recent years, smart contracts running on blockchains often face security vulnerability problems, resulting in significant economic losses. Unlike traditional programs, smart contracts cannot be modified once deployed, and vulnerabilities cannot be remedied. Therefore, the vulnerability detection of smart contracts has become a research focus. Most existing vulnerability detection methods are based on rules defined by experts, which are inefficient and have poor scalability. Although there have been studies using machine learning methods to extract contract features for vulnerability detection, the features considered are singular, and it is impossible to fully utilize smart contract information. In order to overcome the limitations of existing methods, this paper proposes a smart contract vulnerability detection method based on deep learning and multimodal decision fusion. This method also considers the code semantics and control structure information of smart contracts. It integrates the source code, operation code, and control-flow modes through the multimodal decision fusion method. The deep learning method extracts five features used to represent contracts and achieves high accuracy and recall rates. The experimental results show that the detection accuracy of our method for arithmetic vulnerability, re-entrant vulnerability, transaction order dependence, and Ethernet locking vulnerability can reach 91.6%, 90.9%, 94.8%, and 89.5%, respectively, and the detected AUC values can reach 0.834, 0.852, 0.886, and 0.825, respectively. This shows that our method has a good vulnerability detection effect. Furthermore, ablation experiments show that the multimodal decision fusion method contributes significantly to the fusion of different modalities.

VulPathsFinder: A Static Method for Finding Vulnerable Paths in PHP Applications Based on CPG

Today, as PHP application technology is becoming increasingly mature, the functions of modern multi-layer web applications are becoming more and more complete, and the complexity is also gradually increasing. While providing developers with various business functions and interfaces, multi-tier Web applications also successfully cover the details of application development. This type of web application adopts a unified entrance, many object-oriented codes are used, and features such as encapsulation, inheritance, and polymorphism bring challenges to vulnerability mining from the perspective of static analysis. A large amount of object-oriented code makes it impossible for a simple function name-matching method to build a complete call graph (CG), resulting in the inability to perform a comprehensive interprocedural analysis. At the same time, the encapsulation feature of the class makes the data hidden in the object attribute, and the vulnerability path cannot be obtained through the general data-flow analysis. In response to the above issues, we propose a vulnerability detection method that supports vulnerability detection for multi-layer web applications based on MVC (Model-View-Control) architecture. First, we improve the construction of the call graph and Code Property Graph (CPG). Then, based on the enhanced Code Property Graph, we propose a technique to support vulnerability detection for multi-layer web applications. Based on this approach, we implemented a prototype of VulPathsFinder, a security analysis tool extended from the PHP security analyzer Joern-PHP. Then, we select ten MVC based and ten non-MVC-based applications to form a test dataset and validate the effectiveness of VulPathsFinder based on this dataset. Experimental results show that, compared with currently available tools, VulPathsFinder can handle framework applications more effectively, build a complete code property map, and detect vulnerabilities in framework applications that existing tools cannot detect.

Smart contracts vulnerability detection model based on adversarial multi-task learning

Vulnerability detection is important for smart contracts because of their immutable and irreversible features. In this work, a new detection method based on adversarial multi-task learning is proposed to improve the accuracy of existing vulnerability detection methods, which is based on the multi-task learning framework, including a shared part and a task-specific part. We optimize the multi-task learning frameworks and propose the mixed parameter sharing method to make each task not only maintain its uniqueness, but also share features with other tasks, which helps solve the problem that the hard parameter sharing method cannot constrain the underlying shared layer and improve the quality of extracted features. In addition, we introduce adversarial transfer learning to reduce noise pollution caused by the private feature and interference between the general feature and the private feature. We experimented on datasets obtained from our previous work, and the experimental results prove that our proposed model can judge whether there are vulnerabilities in smart contracts and then identify their types. Additionally, the results also show that our model effectively improves detection accuracy and has an advantage in performance over representative methods.

IRC-CLVul: Cross-Programming-Language Vulnerability Detection with Intermediate Representations and Combined Features

The most severe problem in cross-programming languages is feature extraction due to different tokens in different programming languages. To solve this problem, we propose a cross-programming-language vulnerability detection method in this paper, IRC-CLVul, based on intermediate representation and combined features. Specifically, we first converted programs in different programming languages into a unified LLVM intermediate representation (LLVM-IR) to provide a classification basis for different programming languages. Afterwards, we extracted the code sequences and control flow graphs of the samples, used the semantic model to extract the program semantic information and graph structure information, and concatenated them into semantic vectors. Finally, we used Random Forest to learn the concatenated semantic vectors and obtained the classification results. We conducted experiments on 85,811 samples from the Juliet test suite in C, C++, and Java. The results show that our method improved the accuracy by 7% compared with the two baseline algorithms, and the F1 score showed a 12% increase.

Research on smart contract vulnerability detection method based on domain features of solidity contracts and attention mechanism

The smart contract, a self-executing program on the blockchain, is key to programmable finance. However, the rise of smart contract use has also led to an increase in vulnerabilities that attract illegal activity from hackers. Traditional manual approaches for vulnerability detection, relying on domain experts, have limitations such as low automation and weak generalization. In this paper, we propose a deep learning approach that leverages domain-specific features and an attention mechanism to accurately detect vulnerabilities in smart contracts. Our approach reduces the reliance on manual input and enhances generalization by continuously learning code patterns of vulnerabilities, specifically detecting various types of vulnerabilities such as reentrancy, integer overflow, forced Ether injection, unchecked return value, denial of service, access control, short address attack, tx.origin, call stack overflow, timestamp dependency, random number dependency, and transaction order dependency vulnerabilities. In order to extract semantic information, we present a semantic distillation approach for detecting smart contract vulnerabilities. This approach involves using a syntax parser, Slither, to segment the code into smaller slices and word embedding to create a matrix for model training and prediction. Our experiments indicate that the BILSTM model is the best deep learning model for smart contract vulnerability detection task. We looked at how domain features and self-attentiveness mechanisms affected the ability to identify 12 different kinds of smart contract vulnerabilities. Our results show that by including domain features, we significantly increased the F1 values for 8 different types of vulnerabilities, with improvements ranging from 7.35% to 48.58%. The methods suggested in this study demonstrate a significant improvement in F1 scores ranging from 4.18% to 38.70% when compared to conventional detection tools like Oyente, Mythril, Osiris, Slither, Smartcheck, and Securify. This study provides developers with a more effective method of detecting smart contract vulnerabilities, assisting in the prevention of potential financial losses. This research provides developers with a more effective means of detecting smart contract vulnerabilities, thereby helping to prevent potential financial losses.

HGAT: smart contract vulnerability detection method based on hierarchical graph attention network

With the widespread use of blockchain, more and more smart contracts are being deployed, and their internal logic is getting more and more sophisticated. Due to the large false positive rate and low detection accuracy of most current detection methods, which heavily rely on already established detection criteria, certain smart contracts additionally call for human secondary detection, resulting in low detection efficiency. In this study, we propose HGAT, a hierarchical graph attention network-based detection model, in order to address the aforementioned issues as well as the shortcomings of current smart contract vulnerability detection approaches. First, using Syntax Tree (AST) and Control Flow Graph, the functions in the smart contract are abstracted into code graphs (CFG). Then abstract each node in the code subgraph, extract the node features, utilize the graph attention mechanism GAT, splice the obtained vectors to form the features of each line of statements and use these features to detect smart contracts. To create test data and assess HGAT, we leverage the open-source smart contract vulnerability sample dataset. The findings of the experiment indicate that this method can identify smart contract vulnerabilities more quickly and precisely than other detection techniques.

Vulnerability detection through cross-modal feature enhancement and fusion

Software vulnerability detection is critical to computer security. Most existing vulnerability detection methods use single modal-based vulnerability detection models, which cannot effectively extract cross-modal features. To solve this problem, we propose a new multimodal deep learning based vulnerability detection method through a cross-modal feature enhancement and fusion. Firstly, we utilize a special compilation and debugging method to obtain the alignment relationship between source code statements and assembly instructions, as well as between source code variables and assembly code registers. Based on this alignment relationship and program slicing technology, we propose a cross-slicing method to generate bimodal program slices. Then, we propose a cross-modal feature enhanced code representation learning model to capture the fine-grained semantic correlation between source code and assembly code by using the co-attention mechanisms. Finally, vulnerability detection is achieved by feature level fusion of semantic features captured in fine-grained aligned source code and assembly code. Extensive experiments show that our method improves the performance of vulnerability detection compared with state-of-the-art methods. Specifically, our method achieves an accuracy of 97.4% and an F1-measure of 93.4% on the SARD dataset. An average accuracy of 95.4% and an F1-measure of 89.1% on two real-world software projects (i.e., FFmpeg and OpenSSL) is also achieved by our method, improving over SOTA method 4.5% and 2.9%.

Learning and Fusing Multi-View Code Representations for Function Vulnerability Detection

The explosive growth of vulnerabilities poses a significant threat to the security of software systems. While various deep-learning-based vulnerability detection methods have emerged, they primarily rely on semantic features extracted from a single code representation structure, which limits their ability to detect vulnerabilities hidden deep within the code. To address this limitation, we propose S2FVD, short for Sequence and Structure Fusion-based Vulnerability Detector, which fuses vulnerability-indicative features learned from the multiple views of the code for more accurate vulnerability detection. Specifically, S2FVD employs either well-matched or carefully extended neural network models to extract vulnerability-indicative semantic features from the token sequence, attributed control flow graph (ACFG) and abstract syntax tree (AST) representations of a function, respectively. These features capture different perspectives of the code, which are then fused to enable S2FVD to accurately detect vulnerabilities that are well-hidden within a function. The experiments conducted on two large vulnerability datasets demonstrated the superior performance of S2FVD against state-of-the-art approaches, with its accuracy and F1 scores reaching 98.07% and 98.14% respectively in detecting the presence of vulnerabilities, and 97.93% and 97.94%, respectively, in pinpointing specific vulnerability types. Furthermore, with regard to the real-world dataset D2A, S2FVD achieved average performance gains of 6.86% and 14.84% in terms of accuracy and F1 metrics, respectively, over the state-of-the-art baselines. This ablation study also confirms the superiority of fusing the semantics implied in multiple distinct code views to further enhance vulnerability detection performance.

SedSVD: Statement-level software vulnerability detection based on Relational Graph Convolutional Network with subgraph embedding

Current deep-learning based vulnerability detection methods have been proven more automatic and correct to a certain extent, nonetheless, they are limited to detect at function-level or file-level, which can hinder software developers from acquiring more detailed information and conducting more targeted repairs. Graph-based detection methods have shown dominant performance over others. Unfortunately, the information they reveal has not been fully utilized. We design SedSVD (Subgraph embedding driven Statement-level Vulnerability Detection) with two objectives: (i) to better utilize the information the code-related graphs can reflect; (ii) to detect vulnerabilities at a finer-grained level. In our work, we propose a novel graph-based detection framework that embeds graphs at subgraph-level to realize statement-level detection. It first leverages Code Property Graph (CPG) to learn both semantic and syntactic information from source code, and then selects several center nodes (code elements) in CPG to build their subgraphs. After embedding each subgraph with its nodes and edges, we apply Relational Graph Convolutional Network (RGCN) to process different edges differently. A Multi-Layer Perceptron (MLP) layer is further added to ensure its prediction performance. We conduct our experiments on C/C++ projects from NVD and SARD. Experimental results show that SedSVD achieves 95.15% in F1-measure which proves our work to be more effective. Our work detects at a finer-grained level and achieves higher F1-measure than existing state-of-art vulnerability detection techniques. Besides, we provide a more detailed detection report pointing the specific error code elements within statements.

VulGraB: Graph‐embedding‐based code vulnerability detection with bi‐directional gated graph neural network

AbstractCode vulnerabilities can have serious consequences such as system attacks and data leakage, making it crucial to perform code vulnerability detection during the software development phase. Deep learning is an emerging approach for vulnerability detection tasks. Existing deep learning‐based code vulnerability detection methods are usually based on word2vec embedding of linear sequences of source code, followed by code vulnerability detection through RNNs network. However, such methods can only capture the superficial structural or syntactic information of the source code text, which is not suitable for modeling the complex control flow and data flow and miss edge information in the graph structure constructed by the source code, with limited effect of neural network model. To solve the above problems, this article proposes a code vulnerability detection method, named VulGraB, which is based on graph embedding and bidirectional gated graph neural networks. VulGraB uses node2vec to convert the program‐dependent graphs into graph embeddings of the code, which contain rich structure information of the source code, improving the ability of features to express nonlinear information to a certain extent. Then the BiGGNN is used for training, and finally the accuracy of the detection results is evaluated using target program. The bi‐directional gated neural network utilizes a bi‐directional recurrent structure, which is beneficial to global information aggregation. The experimental results show that the accuracy of VulGraB is significantly improved over the baseline models on two datasets, with F1 scores of 85.89% and 97.24% being the highest, demonstrating that VulGraB consistently outperforms other effective vulnerability detection models.