The rise of cryptocurrencies like Bitcoin has not only increased trade volumes but also broadened the use of graph machine learning techniques, such as address embeddings, to analyze transactions and decipher user patterns. Traditional analysis methods rely on simple heuristics and extensive data gathering, while more advanced Graph Neural Networks encounter challenges such as scalability, poor interpretability, and label scarcity in massive blockchain transaction networks. To overcome existing techniques’ computational and interpretability limitations, we introduce a topological approach, Chainlet Orbits, which embeds blockchain addresses by leveraging their topological characteristics in temporal transactions. We employ our innovative address embeddings to investigate financial behavior and e-crime in the Bitcoin and Ethereum networks, focusing on distinctive substructures that arise from user behavior. Our model demonstrates exceptional performance in node classification experiments compared to GNN-based approaches. Furthermore, our approach embeds all daily nodes of the largest blockchain transaction network, Bitcoin, and creates explainable machine learning models in less than 17 minutes which takes days for GNN-based approaches.

The safety and robustness of Large Language Models (LLMs) are major challenges in developing generative AI applications. One key issue is the vulnerability to prompt jailbreak attacks, which pose a significant threat to building secure and resilient LLM-based applications. In this work, we present a framework for understanding and evaluating the behaviors of popular LLMs by categorizing their responses into five distinct exposure levels. Additionally, we introduce a novel language attack that circumvents LLMs’ defenses by translating jailbreak prompts into languages such as Arabic, Chinese, and Greek. Despite ongoing efforts to enhance LLMs’ safety, we find that nearly all popular LLMs can be jailbroken. Our findings offer detailed insights into LLMs’ behavior, improve diagnostic capabilities, and support targeted safety improvements.

The safety and robustness of large language models (LLMs) based applications remain critical challenges in artificial intelligence. Among the key threats to these applications are prompt hacking attacks, which can significantly undermine the security and reliability of LLM-based systems. In this work, we offer a comprehensive and systematic overview of three distinct types of prompt hacking: jailbreaking, leaking, and injection, addressing the nuances that differentiate them despite their overlapping characteristics. To enhance the evaluation of LLM-based applications, we propose a novel framework that categorizes LLM responses into five distinct classes, moving beyond the traditional binary classification. This approach provides more granular insights into the AI’s behavior, improving diagnostic precision and enabling more targeted enhancements to the system’s safety and robustness.

The mainstream adoption of blockchains led to the preparation of many decentralized applications and web platforms, including Web 3.0, a peer-to-peer internet with no single authority. The data stored in blockchain can be considered as big data -- massive-volume, dynamic, and heterogeneous. Due to highly connected structure, graph-based modeling is an optimal tool to analyze the data stored in blockchains. Recently, several research works performed graph analysis on the publicly available blockchain data to reveal insights into its business transactions and for critical downstream tasks, e.g., cryptocurrency price prediction, phishing scams and counterfeit token detection. In this tutorial, we discuss relevant literature on blockchain data structures, storage, categories, data extraction and graphs construction, graph mining, topological data analysis, and machine learning methods used, target applications, and the new insights revealed by them, aiming towards providing a clear view of unified graph-data models for UTXO and account-based blockchains. We also emphasize future research directions.

Blockchain technology garners an ever-increasing interest of researchers in various domains that benefit from scalable cooperation among trust-less parties. As blockchains and their applications proliferate, so do the complexity and volume of data stored by Blockchains. Analyzing this data has emerged as an important research topic, already leading to methodological advancements in information sciences.

Core decomposition in networks has proven useful for evaluating the importance of nodes and communities in a variety of application domains, ranging from biology to social networks and finance. However, existing core decomposition algorithms have limitations in simultaneously handling multiple node and edge attributes.

In recent years, as online social networks continue to grow in size, estimating node features, such as sociodemographics, preferences and health status, in a scalable and reliable way has become a primary research direction in social network mining. Although many techniques have been developed for estimating various node features, quantifying uncertainty in such estimations has received little attention. Furthermore, most existing methods study networks parametrically, which limits insights about necessary quantity of queried data, reliable feature estimation, and estimator uncertainty. Uncertainty quantification is critical for answering key questions, such as, given a limited availability of social network data, how much data should be queried from the network?, and which node features can be learned reliably? More importantly, how can we evaluate uncertainty of our estimators? Uncertainty quantification is not equivalent to network sampling but constitutes a key complementary concept to sampling and the associated reliability analysis. To our knowledge, this paper is the first work that sheds light on uncertainty quantification and uncertainty propagation in social network feature mining. We propose a novel non-parametric bootstrap method for uncertainty analysis of node features in social network mining, derive its asymptotic properties, and demonstrate its effectiveness with extensive experiments. Furthermore, we develop a new metric based on dispersion of estimations, enabling analysts to assess how much more information is needed for increasing prediction reliability based on the estimated uncertainty. We demonstrate the effectiveness of our new uncertainty quantification methodology with extensive experiments on real life social networks, and a case study of mental health on Twitter.

The Blockchain technology and, in particular blockchain-based cryptocurrencies, offer us information that has never been seen before in the financial world. In contrast to fiat currencies, all transactions of crypto-currencies and crypto-tokens are permanently recorded on distributed ledgers and are publicly available. This allows us to construct a transaction graph and to assess not only its organization but to glean relationships between transaction graph properties and crypto price dynamics. The goal of this paper is to facilitate our understanding on horizons and limitations of what can be learned on crypto-tokens from local topology and geometry of the Ethereum transaction network whose even global network properties remain scarcely explored. By introducing novel tools based on Topological Data Analysis and Functional Data Depth into Blockchain Data Analytics, we show that Ethereum network (one of the most popular blockchains for creating new crypto-tokens) can provide critical insights on price changes of crypto-tokens that are otherwise largely inaccessible with conventional data sources and traditional analytic methods.

With emergence of blockchain technologies and the associated cryptocurrencies, such as Bitcoin, understanding network dynamics behind Blockchain graphs has become a rapidly evolving research direction. Unlike other financial networks, such as stock and currency trading, blockchain based cryptocurrencies have the entire transaction graph accessible to the public (i.e., all transactions can be downloaded and analyzed). A natural question is then to ask whether dynamics of the transaction graph impacts price of the underlying cryptocurrency. We show that standard graph features such as degree distribution of the transaction graph may not be sufficient to capture network dynamics and its potential impact on fluctuations of Bitcoin price. In contrast, topological features computed from the blockchain graph using the tools of persistent homology, are found to exhibit higher utility for predicting Bitcoin price dynamics.

Over the last couple of years, Bitcoin cryptocurrency and the Blockchain technology that forms the basis of Bitcoin have witnessed an unprecedented attention. Designed to facilitate a secure distributed platform without central regulation, Blockchain is heralded as a novel paradigm that will be as powerful as Big Data, Cloud Computing, and Machine Learning. The Blockchain technology garners an ever increasing interest of researchers in various domains that benefit from scalable cooperation among trust-less parties. As Blockchain data analytics further proliferates, a need to glean successful approaches and to disseminate them among a diverse body of data scientists became a critical task. As an inter-disciplinary team of researchers, our aim is to fill this vital role. In this tutorial, we offer a holistic view on Blockchain Data Analytics. Starting with the core components of Blockchain, we will discuss the state of art in Blockchain data analytics for privacy, security, finance, and management domains. We will share tutorial notes and further reading pointers on the tutorial website blockchaintutorial.github.io.

Declarative rules, such as functional dependencies, are widely used for cleaning data. Several systems take them as input for detecting errors and computing a "clean" version of the data. To support domain experts, in specifying these rules, several tools have been proposed to profile the data and mine rules. However, existing discovery techniques have traditionally ignored the time dimension. Recurrent events, such as persons reported in locations, have a duration in which they are valid, and this duration should be part of the rules or the cleaning process would simply fail. In this work, we study the rule discovery problem for temporal web data. Such a discovery process is challenging because of the nature of web data; extracted facts are (i) sparse over time, (ii) reported with delays, and (iii) often reported with errors over the values because of inaccurate sources or non robust extractors. We handle these challenges with a new discovery approach that is more robust to noise. Our solution uses machine learning methods, such as association measures and outlier detection, for the discovery of the rules, together with an aggressive repair of the data in the mining step itself. Our experimental evaluation over real-world data from Recorded Future, an intelligence company that monitors over 700K Web sources, shows that temporal rules improve the quality of the data with an increase of the average precision in the cleaning process from 0.37 to 0.84, and a 40% relative increase in the average F-measure.

In the past decade, online social networks have provided invaluable data in understanding how social networks change in time while attracting new users and fostering relationships among members. The concept of social trust was developed to explain why and how much users trust each other to become friends or expose their personal data. Existing work on social trust analyze behavioral features and profile attributes to find trust between pairs of users. Although useful, these works suffer from the problem of incomplete, inaccurate and inconsistent social network data. We approach the problem of analyzing trust from an ego network perspective. We observe new friendships, group formations and structural roles of users in ego networks to outline three trust questions. Answers to these questions provide insights into how social trust can be measured from user connections.

Several efforts have been made for more privacy aware Online Social Networks (OSNs) to protect personal data against various privacy threats. However, despite the relevance of these proposals, we believe there is still the lack of a conceptual model on top of which privacy tools have to be designed. Central to this model should be the concept of risk. Therefore, in this paper, we propose a risk measure for OSNs. The aim is to associate a risk level with social network users in order to provide other users with a measure of how much it might be risky, in terms of disclosure of private information, to have interactions with them. We compute risk levels based on similarity and benefit measures, by also taking into account the user risk attitudes. In particular, we adopt an active learning approach for risk estimation, where user risk attitude is learned from few required user interactions. The risk estimation process discussed in this paper has been developed into a Facebook application and tested on real data. The experiments show the effectiveness of our proposal.

In this paper, we explore the risks of friends in social networks caused by their friendship patterns, by using real life social network data and starting from a previously defined risk model. Particularly, we observe that risks of friendships can be mined by analyzing users' attitude towards friends of friends. This allows us to give new insights into friendship and risk dynamics on social networks.