Selin Sezer, Clemens Eyhoff, Wolfgang Prinz, Thomas Rose
Due to the increase in smart contracts in Ethereum, a need for proper classification has emerged. Although the smart contracts are accessible due to the open nature of the Blockchain, readability is still an issue with respect to the smart contract bytecode. We propose an automated approach for classifying smart contracts that utilize popular text classification methods on the opcode translation of the smart contract bytecode in order to overcome this limitation. Our experiments indicate that the decision-tree-based techniques like Random Forest and Xgboost outmatch the traditional classification tools like Naïve Bayes, Logistic Regression, and SVM once the opcode input is presented as n-gram tf-idf vectors.
Gang Tian, Qibo Wang, Yi Zhao, Lantian Guo · 6 authors
With the number of smart contracts growing rapidly, retrieving the relevant smart contracts quickly and accurately has become an important issue. A key step for recognizing the related smart contracts is able to classify them accurately. Different from traditional text, the smart contract is composed of several parts: source code, code comments and other useful information like account information. How to make good use of those different kinds of features for effective classification is a problem need to be solved. Inspired by this, we proposed a smart contract classification approach based on Bi-LSTM model and Gaussian LDA, which can use a variety of information as inputs of the model, including source code, comments, tags, account and other content information. Bi-LSTM is utilized to capture grammar rules and context information in source code, while Gaussian LDA model is employed to generate comments feature where the semantics of the comments are enriched by embeddings. We also use attention mechanism to focus on the more relevant features in smart contracts for tags and fuse account information to provide additional information for classification. The experimental results show that the classification performance of the proposed model is superior to other baseline models.
Evidence and witness play an important role to investigate crime and lawful jurisdiction in any case. But often, the victims do not get justice due to the middlemen and the altered evidences in the centralized network. Block chain is an ideal solution to ensure transparency till the highest hierarchy of jurisdiction using a decentralized peer to peer network to store data which is immutable. It is like a distributed ledger working on the proof of work algorithm that validates each amendment and modification made in a particular chain. Evichain inculcates the transparency and security of the same decentralized network in the crime investigation process. It is an application on which the entire data regarding any particular investigation is stored in a block chain with limited people having access rights. It also includes cases filed by the victims themselves. Every amendment made in the block chain is validated. Also the data access is restricted to the users with that specific private key.
This research work proposes a novel and comprehensive electronic cheque transactions framework. The proposed e-cheque system is free from the various security attacks such as alteration of the e-cheque, double spending of e-cheque, counterfeits e-cheques. The e-cheque generated in the proposed system can be deposited electronically or physically via teller machines. This facility provides greater flexibility to the customers of the banking system. The proposed system also provides space for professional miners to participate in the e-cheque transaction system by performing mining's and earn incentives. As the customer's perspective of security, the proposed system uses digital signature and cryptographic hash in each transaction hence it is a completely secure system. The existing CTS based cheque clearance request requires at least one day to clear a cheque which could extend to two or three days but the proposed system requires only 1.65 seconds for clearing any e-cheque.
Bitcoin is a cryptocurrency that features a distributed, decentralized and trustworthy mechanism, which has made Bitcoin a popular global transaction platform. The transaction efficiency among nations and the privacy benefiting from address anonymity of the Bitcoin network have attracted many activities such as payments, investments, gambling, and even money laundering in the past decade. Unfortunately, some criminal behaviors which took advantage of this platform were not identified. This has discouraged many governments to support cryptocurrency. Thus, the capability to identify criminal addresses becomes an important issue in the cryptocurrency network. In this paper, we propose new features in addition to those commonly used in the literature to build a classification model for detecting abnormality of Bitcoin network addresses. These features include various high orders of moments of transaction time (represented by block height) which summarizes the transaction history in an efficient way. The extracted features are trained by supervised machine learning methods on a labeling category data set. The experimental evaluation shows that these features have improved the performance of Bitcoin address classification significantly. We evaluate the results under eight classifiers and achieve the highest Micro-Fl /Macro-F1 of 87% /86% with LightGBM.
Blockchain technology becomes increasingly popular. It also attracts scams, for example, a Ponzi scheme, a classic fraud, has been found making a notable amount of money on Blockchain, which has a very negative impact. To help to deal with this issue and to provide reusable research data sets for future research, this paper collects real-world samples and proposes an approach to detect Ponzi schemes implemented as smart contracts (i.e., smart Ponzi schemes) on the blockchain. First, 200 smart Ponzi schemes are obtained by manually checking more than 3,000 open source smart contracts on the Ethereum platform. Then, two kinds of features are extracted from the transaction history and operation codes of the smart contracts. Finally, a classification model is presented to detect smart Ponzi schemes. The extensive experiments show that the proposed model performs better than many traditional classification models and can achieve high accuracy for practical use. By using the proposed approach, we estimate that there are more than 500 smart Ponzi schemes running on Ethereum. Based on these results, we propose to build a uniform platform to evaluate and monitor every created smart contract for early warning of scams.
SUMMARY A perennial challenge in the accounting profession is how to aggregate and share instances of practitioner misconduct among numerous relevant parties. At present, both the American Institute of Certified Public Accountants (AICPA) and National Association of State Boards of Accountancy (NASBA) offer solutions for centralized collection of misconduct, but both likely experience issues with incomplete reporting from key constituents. I propose a novel use of blockchain technology to address this issue, such that all key parties in the accounting profession leverage an accountancy blockchain to aggregate and share instances of practitioner misconduct across the country on a nearly real-time basis. Such a network creates an immutable record of misconduct and allows key constituents in the accounting profession to work together and share information as peers without the risk of one party taking control of the ledger. I close by discussing blockchain-specific roadblocks to realizing this proposed model.
Blockchain technology is already being talked about as one of the megatrends for the next years. Researchers and organisations are starting to understand the potential benefits of this technology and are exploring how it can disrupt the world we live in with a diverse range of applications. But the truth is the ability to move blockchain from concept to adoption and production has been minimal yet. When it comes to auditing, blockchain solutions could have important benefits by reducing the workload of the auditors, helping in minimising fraud and optimising the existing processes but is also vital to have in mind other emerging technologies. Factom, Libra, and Verady are some examples of companies developing blockchain solutions that can be applied in the auditing environment, but much of the necessary development is still yet to be done.
Both “big data” and “analytics” have become popular keywords in many organizations. The power data analytics has on harnessing the increasing volumes, velocity and complexity of data in a world of constant change and disruptive technologies has been recognized. Many companies are making significant investments to better understand the impact of these capabilities on their businesses. One area with significant potential is the transformation of the audit. This project explores ways in which analytics can change and shape the work of accountants. \n \nAnomaly detection plays a pivotal role in data mining since most outlying points contain crucial information for further investigation. In the financial world which the Bitcoin network is a part of, anomaly detection can indicate fraud. Using data mining tools such as Regression, we simultaneously examine the relationship among variables whilst visually inspecting the data for possible outliers. By doing so, I have chosen the world’s leading cryptocurrency, Bitcoin. This project will conclude with an in-depth analysis on whether or not data analytics can shape how effectively, and secure accountants can audit transactions by implementing analytics tools into their daily protocols.
This dissertation consists of three essays that design and evaluate the continuous audit analytics and fraud prevention systems using three emerging technologies (i.e., the blockchain, in-memory cloud computing, and deep learning). The first essay designs a framework of Blockchain-based Transaction Processing System using the homomorphic encryption and zero-knowledge proof mechanisms. Furthermore, this study develops a prototype of the designed system to demonstrate its applications in real-time accounting, continuous monitoring, and fraud prevention. Although the simulation tests show the Blockchain-based Transaction Processing System consumes more computational overhead than the conventional database-based ERP system, the blockchain should be considered as a promising technology for future accounting and auditing practice. The second essay introduces the database architecture that manages data in main physical memory and columnar format. This essay proposes a conceptual framework for applying the in-memory columnar database system to support high-speed continuous audit analytics. Moreover, this study develops a prototype and conducts the simulation tests to evaluate the proposed framework. The test results show the high efficiency and effectiveness of the in-memory columnar database relative to the conventional ERP system regarding the computational time and the storage volume. Furthermore, the deployment of the in-memory columnar database to the cloud shows great promise of applying the in-memory columnar database for continuous audit analytics. The third essay designs a continuous fraud detection system based on modified deep learning technology. Specifically, this essay builds an accounting layer on top of the deep learning architecture to process financial data for predicting the fraudulent financial statements. A prototype is developed to evaluate the prediction accuracy of the proposed design. The test results show the deep learning-based continuous fraud detection system provides high prediction accuracy relative to the existing studies of financial statement fraud detection.
This paper addresses recognizing fraud users on a Bitcoin exchange website-bitcoin-otc. According to online rating records provided by the website, some users behave significantly different from others. Seeing that, the classical K-means clustering algorithm is proposed to identify these abnormal users. K-means algorithm is an unsupervised clustering algorithm that clusters users based on feature similarity. Therefore, performance of K-means algorithm relies on the features. This paper explored and found the best collection of features based on real record data, e.g., mean of total ratings sent. Since the selected features are not observed for record set, the website should offer these features for potential traders.
Bitcoin is the first decentralized peer-to-peer payment network that is powered by its users with no central authority or middlemen. The objective of this study is to evaluate the normality of data distribution for exchange rate of Bitcoin. The method implemented in this study is Shapiro-Wilk normality test including graphical approach namely box plot .Results show the data distribution of exchange rate for Bitcoin follows non-normal distribution. Therefore, the normality transformation is important to make sure the distribution of data follows normal distribution. The normal distribution is very crucial as one of the requirement for validity of statistical test.Normality tests are used to determine if a data set is well-modeled by a normal distribution and to compute how likely it is for a random variable underlying the data set to be normally distributed.This study implemented two-stages of outliers detection and deletion process.The final results shows the distribution of Bitcoin exchange rate with first difference is follow normal distribution with probability of 0.722.Result concluded the distribution of data after second stages of outlies deletion treatment shows high normal distribution characteristics. This finding concludes that Bitcoin data is highly volatile with existence of many outliers. The transformation process is highly important to make sure the Bitcoin data follows normal distribution that underlying critical assumption for statistical tests.
The problem of anomaly detection has been studied for a long time. In short, anomalies are abnormal or unlikely things. In financial networks, thieves and illegal activities are often anomalous in nature. Members of a network want to detect anomalies as soon as possible to prevent them from harming the network's community and integrity. Many Machine Learning techniques have been proposed to deal with this problem; some results appear to be quite promising but there is no obvious superior method. In this paper, we consider anomaly detection particular to the Bitcoin transaction network. Our goal is to detect which users and transactions are the most suspicious; in this case, anomalous behavior is a proxy for suspicious behavior. To this end, we use three unsupervised learning methods including k-means clustering, Mahalanobis distance, and Unsupervised Support Vector Machine (SVM) on two graphs generated by the Bitcoin transaction network: one graph has users as nodes, and the other has transactions as nodes.