Estimating causal effects in e-commerce tends to involve costly treatment assignments which can be impractical in large-scale settings. Leveraging machine learning to predict such treatment effects without actual intervention is a standard practice to diminish the risk. However, existing methods for treatment effect prediction tend to rely on training sets of substantial size, which are built from real experiments and are thus inherently risky to create. In this work we propose a graph neural network to diminish the required training set size, relying on graphs that are common in e-commerce data. Specifically, we view the problem as node regression with a restricted number of labeled instances, develop a two-model neural architecture akin to previous causal effect estimators, and test varying message-passing layers for encoding. Furthermore, as an extra step, we combine the model with an acquisition function to guide the creation of the training set in settings with extremely low experimental budget. The framework is flexible since each step can be used separately with other models or policies. The experiments on real large-scale networks indicate a clear advantage of our methodology over the state of the art, which in many cases performs close to random underlining the need for models that can generalize with limited labeled samples to reduce experimental risks.
In multiagent systems (MASs), agents' observation upon system behaviours may improve the overall team performance, but may also leak sensitive information to an observer. A quantified observability analysis can thus be useful to assist decision-making in MASs by operators seeking to optimise the relationship between performance effectiveness and information exposure through observations in practice. This paper presents a novel approach to quantitatively analysing the observability properties in MASs. The concept of opacity is applied to formally express the characterisation of observability in MASs modelled as partially observable multiagent systems. We propose a temporal logic oPATL to reason about agents' observability with quantitative goals, which capture the probability of information transparency of system behaviours to an observer, and develop verification techniques for quantitatively analysing such properties. We implement the approach as an extension of the PRISM model checker, and illustrate its applicability via several examples.
Generative adversarial networks (GANs) have shown remarkable success in image synthesis, making GAN models themselves commercially valuable to legitimate model owners. Therefore, it is critical to technically protect the intellectual property of GANs. Prior works need to tamper with the training set or training process, and they are not robust to emerging model extraction attacks. In this paper, we propose a new ownership protection method based on the common characteristics of a target model and its stolen models. Our method can be directly applicable to all well-trained GANs as it does not require retraining target models. Extensive experimental results show that our new method can achieve the best protection performance, compared to the state-of-the-art methods. Finally, we demonstrate the effectiveness of our method with respect to the number of generations of model extraction attacks, the number of generated samples, different datasets, as well as adaptive attacks.
Diffusion models have been remarkably successful in data synthesis. Such successes have also driven diffusion models to apply to sensitive data, such as human face data, but this might bring about severe privacy concerns. In this work, we systematically present the first privacy study about property inference attacks against diffusion models, in which adversaries aim to extract sensitive global properties of the training set from a diffusion model, such as the proportion of the training data for certain sensitive properties. Specifically, we consider the most practical attack scenario: adversaries are only allowed to obtain synthetic data. Under this realistic scenario, we evaluate the property inference attacks on different types of samplers and diffusion models. A broad range of evaluations shows that various diffusion models and their samplers are all vulnerable to property inference attacks. Furthermore, one case study on off-the-shelf pre-trained diffusion models also demonstrates the effectiveness of the attack in practice. Finally, we propose a new model-agnostic plug-in method PriSampler to mitigate the property inference of diffusion models. PriSampler can be directly applied to well-trained diffusion models and support both stochastic and deterministic sampling. Extensive experiments illustrate the effectiveness of our defense and it makes adversaries infer the proportion of properties as close as random guesses. PriSampler also shows its significantly superior performance to diffusion models trained with differential privacy on both model utility and defense performance.
Recent years have witnessed the tremendous success of diffusion models in data synthesis. However, when diffusion models are applied to sensitive data, they also give rise to severe privacy concerns. In this paper, we systematically present the first study about membership inference attacks against diffusion models, which aims to infer whether a sample was used to train the model. Two attack methods are proposed, namely loss-based and likelihood-based attacks. Our attack methods are evaluated on several state-of-the-art diffusion models, over different datasets in relation to privacy-sensitive data. Extensive experimental evaluations show that our attacks can achieve remarkable performance. Furthermore, we exhaustively investigate various factors which can affect attack performance. Finally, we also evaluate the performance of our attack methods on diffusion models trained with differential privacy.
To address the vaccine hesitancy which impairs the efforts of the COVID-19 vaccination campaign, it is imperative to understand public vaccination attitudes and timely grasp their changes. In spite of reliability and trustworthiness, conventional attitude collection based on surveys is time-consuming and expensive, and cannot follow the fast evolution of vaccination attitudes. We leverage the textual posts on social media to extract and track users' vaccination stances in near real time by proposing a deep learning framework. To address the impact of linguistic features such as sarcasm and irony commonly used in vaccine-related discourses, we integrate into the framework the recent posts of a user's social network neighbours to help detect the user's genuine attitude. Based on our annotated dataset from Twitter, the models instantiated from our framework can increase the performance of attitude extraction by up to 23% compared to state-of-the-art text-only models. Using this framework, we successfully validate the feasibility of using social media to track the evolution of vaccination attitudes in real life. We further show one practical use of our framework by validating the possibility to forecast a user's vaccine hesitancy changes with information perceived from social media.
Vaccine hesitancy is considered as one main cause of the stagnant uptake ratio of COVID-19 vaccines in Europe and the US where vaccines are sufficiently supplied. Fast and accurate grasp of public attitudes toward vaccination is critical to address vaccine hesitancy, and social media platforms have proved to be an effective source of public opinions. In this paper, we describe the collection and release of a dataset of tweets related to COVID-19 vaccines. This dataset consists of the IDs of 2,198,090 tweets collected from Western Europe, 17,934 of which are annotated with the originators' vaccination stances. Our annotation will facilitate using and developing data-driven models to extract vaccination attitudes from social media posts and thus further confirm the power of social media in public health surveillance. To lay the groundwork for future research, we not only perform statistical analysis and visualisation of our dataset, but also evaluate and compare the performance of established text-based benchmarks in vaccination stance extraction. We demonstrate one potential use of our data in practice in tracking the temporal changes of public COVID-19 vaccination attitudes.
Graph Neural Networks (GNNs) have been predominant for graph learning tasks; however, recent studies showed that a well-known graph algorithm, Label Propagation (LP), combined with a shallow neural network can achieve comparable performance to GNNs in semi-supervised node classification on graphs with high homophily. In this paper, we show that this approach falls short on graphs with low homophily, where nodes often connect to the nodes of the opposite classes. To overcome this, we carefully design a combination of a base predictor with LP algorithm that enjoys a closed-form solution as well as convergence guarantees. Our algorithm first learns the class compatibility matrix and then aggregates label predictions using LP algorithm weighted by class compatibilities. On a wide variety of benchmarks, we show that our approach achieves the leading performance on graphs with various levels of homophily. Meanwhile, it has orders of magnitude fewer parameters and requires less execution time. Empirical evaluations demonstrate that simple adaptations of LP can be competitive in semi-supervised node classification in both homophily and heterophily regimes.
Recently, supervised network embedding (NE) has emerged as a predominant technique for representing complex systems that take the form of networks, and various downstream node- and network-level tasks have benefited from its remarkable developments. However, unsupervised NE still remains challenging due to the uncertainty in defining a learning objective. In addition, it is still an unexplored research question whether existing NE methods adapt well to heterophilous networks. This paper introduces the first empirical study on the influence of homophily ratio on the performance of existing unsupervised NE methods and reveals their limitations. Inspired by our empirical findings, we design unsupervised NE task as an r-ego network discrimination problem and further develop a SELf-supErvised Network Embedding (Selene) framework for learning useful node representations for both homophilous and heterophilous networks. Specifically, we propose a dual-channel feature embedding mechanism to fuse node attributes and network structure information and leverage a sampling and anonymisation strategy to break the implicit homophily assumption of existing embedding mechanisms. Lastly, we introduce a negative-sample-free SSL objective function to optimise the framework. We conduct extensive experiments and a series of ablation studies on 12 real-world datasets and 20 synthetic networks. Results demonstrate Selene's superior performance and confirm the effectiveness of each component.
Model extraction attacks aim to duplicate a machine learning model through query access to a target model. Early studies mainly focus on discriminative models. Despite the success, model extraction attacks against generative models are less well explored. In this paper, we systematically study the feasibility of model extraction attacks against generative adversarial networks (GANs). Specifically, we first define accuracy and fidelity on model extraction attacks against GANs. Then we study model extraction attacks against GANs from the perspective of accuracy extraction and fidelity extraction, according to the adversary's goals and background knowledge. We further conduct a case study where an adversary can transfer knowledge of the extracted model which steals a state-of-the-art GAN trained with more than 3 million images to new domains to broaden the scope of applications of model extraction attacks. Finally, we propose effective defense techniques to safeguard GANs, considering a trade-off between the utility and security of GAN models.