WWW 2025 Papers

The rapid growth of decentralized applications, while revolutionizing financial transactions, has created an attractive target for malicious attacks. Existing approaches to detecting attacks often rely on predefined rules or simplistic and overly-specialized models, which lack the flexibility to handle the wide spectrum of diverse and dynamically changing attack types. To address this challenge, we present a general, extensible framework, MoE (Monitoring Ethereum), that leverages runtime verification to detect a wide range of attacks on Ethereum. MoE features an expressive attack modeling language, based on Metric First-order Temporal Logic, that can formalize a wide range of attacks. We integrate a novel semantic lifting approach that extracts vital system behaviors for various attacks utilizing the monitoring tool MonPoly. We further equip MoE with quantitative capabilities to evaluate the similarity between a transaction and an attack formula to identify more attacks, including near-miss attacks. We carry out extensive experiments with MoE on a labeled benchmark and a large-scale dataset containing over one million transactions. On the labeled benchmark, MoE successfully detects 92.0% attacks and achieves 45.0% more recall rate than another state-of-the-art tool. MoE finds 3,319 attacks with 95.4% precision on the large dataset. Furthermore, MoE uses quantitative analysis to uncover 8% more attacks. Notably, the average time for monitoring a transaction is less than 23 ms, positioning MoE as a promising practical solution for real-time attack detection for Ethereum.

Graph-based approximate nearest neighbor search (ANNS) algorithms are widely used to identify the most similar vectors to a given query vector. Graph-based ANNS consists of two stages: constructing a graph and searching on the graph for a given query vector. While reducing the query response time is of great practical importance, less attention has been paid to improving the online search method than the offline graph construction method. This paper provides an extensive experimental analysis on the popular greedy search and other search optimization strategies. We also propose a novel angular distance-guided search method for graph-based ANNS (ADA-NNS) to improve search efficiency. The key innovation of ADA-NNS is introducing a low-cost neighbor selection mechanism based on approximate similarity score derived from angular distance estimation, which effectively filters out less relevant neighbors. We compare state-of-the-art search techniques, including FINGER, on six datasets using different similarity metrics. It provides a comprehensive perspective on their tradeoffs in terms of throughput, latency, and recall. Our evaluation shows that ADA-NNS achieves 34%-107% higher queries per second (QPS) than the greedy search at 95% recall@10 on HNSW, one of the most popular graph structures for ANNS.

With the rapid development of Large Language Models (LLMs), recent studies employed LLMs as recommenders to provide personalized information services for distinct users. Despite efforts to improve the accuracy of LLM-based recommendation models, relatively little attention is paid to beyond-utility dimensions. Moreover, there are unique evaluation aspects of LLM-based recommendation models, which have been largely ignored. To bridge this gap, we explore four new evaluation dimensions and propose a multidimensional evaluation framework. The new evaluation dimensions include: 1) history length sensitivity, 2) candidate position bias, 3) generation-involved performance, and 4) hallucinations. All four dimensions have the potential to impact performance, but are largely unnecessary for consideration in traditional systems. Using this multidimensional evaluation framework, along with traditional aspects, we evaluate the performance of seven LLM-based recommenders, with three prompting strategies, comparing them with six traditional models on both ranking and re-ranking tasks on four datasets. We find that LLMs excel at handling tasks with prior knowledge and shorter input histories in the ranking setting, and perform better in the re-ranking setting, beating traditional models across multiple dimensions. However, LLMs exhibit substantial candidate position bias issues, and some models hallucinate non-existent items much more often than others. We intend our evaluation framework and observations to benefit future research on the use of LLMs as recommenders. The code and data are available at https://anonymous.4open.science/r/EvaLLMasRecommender-3118/.

In the literature, prior studies on Video Anomaly Detection (VAD) mainly focus on detecting whether each video frame is abnormal or not in the video, which largely ignore the structured video semantic information (i.e., what, when, and where does the abnormal event happen), though this structured information could be employed to construct a more precise and efficient system for abnormal event monitoring and retrieval. With this in mind, we propose a new chat-paradigm Multi-scene Video Abnormal Event Extraction and Localization (M-VAE) task, aiming to extract the abnormal event quadruples (i.e., subject, event type, object, scene) and localize such event. Further, this paper believes that this new task faces two key challenges, i.e., global-local spatial modeling and global-local spatial balancing. To this end, this paper proposes a Global-local Spatial-sensitive Large Language Model (LLM) named Sherlock, i.e., acting like Sherlock Holmes to track down the criminal events, for this M-VAE task. Specifically, this approach designs a Global-local Spatial-enhanced MoE (GSM) module and a Spatial Imbalance Regulator (SIR) to address the above two challenges respectively. Extensive experiments on our constructed M-VAE instruction dataset show the significant advantages of Sherlock over several advanced Video-LLMs. This justifies the importance of global-local spatial information for the M-VAE task and the effectiveness of Sherlock in capturing such information.

Recently, many e-commerce platforms have favored presenting a mixed list of ads and organic content to users. The widely-used approach separately ranks ads and organic items, then sequentially inserts ads into the list of organic items. However, this method yields sub-optimal results. Firstly, it only ensures that each generated ad and organic item list achieves local optimality, while the predetermined insertion order fails to guarantee global optimality. Secondly, this approach overlooks the mutual effect between organic items and ads, resulting in an incomplete utilization of contextual information. Besides, it cannot prevent strategic behavior by advertisers. Therefore, we propose a context-aware integrated framework to address these issues. This framework applies automated mechanism design to integrated ad auctions for the first time. Specifically, it models ads and organic items simultaneously along with their contextual information and employs a learning-based approach to prevent advertisers from engaging in strategic behavior. Afterward, the framework directly generates a mixed list, enhancing the overall performance. We also propose $\textbf{T}$ransformer encoder-based $\textbf{I}$ntegrated $\textbf{C}$ontextual $\textbf{Net}$work (TICNet) to generate the optimal integrated contextual ad auction. Finally, we validate the effectiveness of TICNet on synthetic and real-world datasets. Our experimental results demonstrate that TICNet significantly outperforms baseline models across multiple metrics.

Machine learning (ML) based malware classifiers are widely deployed in web applications. Training such classifiers often relies on crowdsourced threat feeds, creating a natural attack point. Recent studies show that attackers can misguide models by injecting trigger embedded samples during training. In the malware domain, attackers are typically limited to clean-label attacks, where they lack control over data labeling. However, clean-label attacks often suffer from suboptimal performance due to competition between trigger features and original clean features during training. Existing studies typically construct poisoned samples by embedding triggers into randomly selected benignware (a method referred to as "random selection"). However, not all benignware are equally suitable for trigger embedding, as the degree of competition between trigger features and original clean features may vary among different benignware. To enhance the effectiveness of clean-label attacks, we propose a simple yet effective sample selection method, called $\textbf{P}$oisoning $\textbf{M}$alware-$\textbf{S}$imilar $\textbf{B}$enignware $\textbf{(PMSB)}$, to identify samples to be poisoned. It reduces the competition between trigger features and original clean features during model training, thereby enhancing the influence of trigger features on the model's decision-making. Additionally, to identify malware-similar benignware, we introduce three distance metrics from different perspectives for sample selection, allowing it to adapt to varying data distributions. Extensive evaluations on three datasets under different attack settings demonstrate the superiority and broad applicability of PMSB, achieving an improvement in attack success rate of over 23.97%.

Recommender systems need to optimize various types of user feedback, e.g., clicks, likes, and shares. A typical recommender system handling multiple types of feedback has two components: a multi-task learning (MTL) module, predicting feedback such as click-through rate and like rate; and a multi-task fusion (MTF) module, integrating these predictions into a single score for item ranking. MTF is essential for ensuring user satisfaction, as it directly influences recommendation outcomes. Recently, reinforcement learning (RL) has been applied to MTF tasks to improve long-term user satisfaction. However, existing RL-based MTF methods are formula-based methods, which only adjust limited coefficients within pre-defined formulas. The pre-defined formulas restrict the RL search space and become a bottleneck for MTF. To overcome this, we propose a formula-free MTF framework. We demonstrate that any suitable fusion function can be expressed as a composition of single-variable monotonic functions, as per the Sprecher Representation Theorem. Leveraging this, we introduce a novel learnable monotonic fusion cell (MFC) to replace pre-defined formulas. We call this new MFC-based model eXtreme MTF (xMTF). Furthermore, we employ a two-stage hybrid (TSH) learning strategy to train xMTF effectively. By expanding the MTF search space, xMTF outperforms existing methods in extensive offline and online experiments. xMTF has been deployed online, serving over 100 million users.

Representative subset selection from data streams is a critical problem with wide-ranging applications in web data mining and machine learning, such as social media marketing, big data summarization, and recommendation systems. This problem is often framed as maximizing a monotone submodular function subject to a knapsack constraint, where each data element in the stream has an associated cost, and the goal is to select elements within a budget $B$ to maximize revenue. However, existing algorithms typically rely on restrictive assumptions about the costs of data elements, and their performance bounds heavily depend on the budget $B$. As a result, these algorithms are only effective in limited scenarios and have super-linear time complexity, making them unsuitable for large-scale data streams. In this paper, we introduce the first linear-time streaming algorithms for this problem, without any assumptions on the data stream, while also minimizing memory usage. Specifically, our single-pass streaming algorithm achieves an approximation ratio of $1/8-\epsilon$ under $\mathcal{O}(n)$ time complexity and $\mathcal{O}(k\log\frac{1}{\epsilon})$ space complexity, where $k$ is the largest cardinality of any feasible solution. Our multi-pass streaming algorithm improves this to a $(1/2-\epsilon)$-approximation using only three passes over the stream, with $\mathcal{O}(\frac{n}{\epsilon}\log\frac{1}{\epsilon})$ time complexity and $\mathcal{O}(\frac{k}{\epsilon}\log\frac{1}{\epsilon})$ space complexity. Extensive experiments across various applications related to web data mining and social media marketing demonstrate the superiority of our algorithms in terms of both effectiveness and efficiency.

Network Intrusion Detection Systems (NIDS) are critical for web security by identifying and blocking malicious traffic. In-network NIDS leverage programmable switches for high-speed traffic processing. However, they are unable to reconcile the fine-grained classification of known classes and the identification of unseen attacks. Moreover, they lack support for incremental updates. In this paper, we propose Helios, an in-network malicious traffic detection system, for continual adaptation in attack-incremental scenarios. First, we design a novel Supervised Mixture Prototypical Learning (SMPL) method combined with clustering initialization to learn prototypes that encapsulate the knowledge, based on the weighted infinity norm distance. SMPL enables known class classification and unseen attack identification through similarity comparison between prototypes and samples. Then, we design boundary calibration and overlap refinement to transform learned prototypes into priority-guided matching rules, ensuring precise and efficient in-network deployment. Additionally, Helios supports incremental prototype learning and rule updates, achieving low-cost hardware reconfiguration. We implement Helios on a Tofino switch and evaluation on three datasets shows that Helios achieves superior performance in classifying known classes (92\%+ in ACC and F1) as well as identifying unseen attacks (62\% - 98\% in TPR). Helios has also reduced resource consumption and reconfiguration time, demonstrating its scalability and efficiency for real-world deployment.

Graph neural networks (GNNs) exhibit powerful performance in handling graph data, with Euclidean and hyperbolic variants excelling in processing grid-based and hierarchical structures, respectively. However, existing methods focus on learning specific structures that are linked to the inherent properties of the underlying space, and fail to fully exploit their complementary properties in distinct geometric spaces, thereby limiting their ability to efficiently model and represent complex graph structures. In this paper, we propose a Hyperbolic-Euclidean Deep Mutual Learning (H-EDML) framework, which leverages the unique properties of hyperbolic space to effectively capture the hierarchical relationships present in graph data, while also utilizes the familiar Euclidean space to handle local interactions. Specifically, We design a topology mutual learning module to bolster the capacity of each single model to perceive the holistic topological structure of the graph. Then, we integrate a decision mutual learning module to further advance the models' comprehensive judgment capabilities towards graph data, thereby strengthening the robustness and generalization. Furthermore, we employ an attention-based probabilistic integration strategy for the final prediction to alleviate potential disparities in decision-making among different models. Extensive experiments on node classification are conducted on five real-world graph datasets and the results show that our proposed H-EDML achieves competitive performances compared to the state-of-the-art methods.

Recent research on explainable recommendation generally frames the task as a standard text generation problem, and evaluates models simply based on the textual similarity between the predicted and ground-truth explanations. However, this approach fails to consider one crucial aspect of the systems: whether their outputs accurately reflect the users' (post-purchase) sentiments, i.e., whether and why they would like and/or dislike the recommended items. To shed light on this issue, we introduce new datasets and evaluation methods that focus on the users' sentiments. Specifically, we construct the datasets by explicitly extracting users' positive and negative opinions from their post-purchase reviews using an LLM, and propose to evaluate systems based on whether the generated explanations 1) align well with the users' sentiments, and 2) accurately identify both positive and negative opinions of users on the target items. We benchmark several recent models on our datasets and demonstrate that achieving strong performance on existing metrics does not ensure that the generated explanations align well with the users' sentiments. Lastly, we find that existing models can provide more sentiment-aware explanations when the users' (predicted) ratings for the target items are directly fed into the models as input. We will release our code and datasets upon acceptance.

Signed graph clustering is a critical technique for discovering community structures in graphs that exhibit both positive and negative relationships. We have identified two significant challenges in this domain: i) existing signed spectral methods are highly vulnerable to noise, which is prevalent in real-world scenarios; ii) the guiding principle ``an enemy of my enemy is my friend'', rooted in \textit{Social Balance Theory}, often narrows or disrupts cluster boundaries in mainstream signed graph neural networks. Addressing these challenges, we propose the \underline{D}eep \underline{S}igned \underline{G}raph \underline{C}lustering framework (DSGC), which leverages \textit{Weak Balance Theory} to enhance preprocessing and encoding for robust representation learning. First, DSGC introduces Violation Sign-Refine to denoise the signed network by correcting noisy edges with high-order neighbor information. Subsequently, Density-based Augmentation enhances semantic structures by adding positive edges within clusters and negative edges across clusters, following \textit{Weak Balance} principles. The framework then utilizes \textit{Weak Balance} principles to develop clustering-oriented signed neural networks to broaden cluster boundaries by emphasizing distinctions between negatively linked nodes. Finally, DSGC optimizes clustering assignments by minimizing a regularized clustering loss. Comprehensive experiments on synthetic and real-world datasets demonstrate DSGC consistently outperforms all baselines, establishing a new benchmark in signed graph clustering. The code is provided in https://anonymous.4open.science/r/DSGC-C05C/.

Web refresh crawling is the problem of keeping a cache of web pages fresh, that is, having the most recent copy available when a page is requested, given a limited bandwidth available to the crawler. Under the assumption that the change and request events, resp., to each web page follow independent Poisson processes, the optimal scheduling policy was derived by Azar et. al (2018). In this paper, we study an extension of this problem where side information indicating content changes, such as various types of web pings, e.g., signals from sitemaps, content delivery networks, etc., is available. Incorporating such side information into the crawling policy is challenging, because (i) the signals can be noisy with false positive events and with missing change events; and (ii) the crawler should achieve a fair performance over web pages regardless of the quality of the side information, which might differ from web page to web page. We propose a scalable crawling algorithm which (i) uses the noisy side information in an optimal way under mild assumptions; (ii) can be deployed without heavy centralized computation; (iii) is able to crawl web pages at a constant total rate without spikes in the total bandwidth usage over any time interval, and automatically adapt to the new optimal solution when the total bandwidth changes without centralized computation. Experiments clearly demonstrate the versatility of our approach.

While implicit feedback is foundational to modern recommender systems, factors such as human error, uncertainty, and ambiguity in user behavior inevitably introduce significant noise into this feedback, adversely affecting the accuracy and robustness of recommendations. To address this issue, existing methods typically aim to reduce the training weight of noisy feedback or discard it entirely, based on the observation that noisy interactions often exhibit higher losses in the overall loss distribution. However, we identify two key issues: (1) there is a significant overlap between normal and noisy interactions in the overall loss distribution, and (2) this overlap becomes even more pronounced when transitioning from pointwise loss functions (e.g., BCE loss) to pairwise loss functions (e.g., BPR loss). This overlap leads traditional methods to misclassify noisy interactions as normal, and vice versa. To tackle these challenges, we further investigate the loss overlap and find that for a given user, there is a clear distinction between normal and noisy interactions in the user's personal loss distribution. Based on this insight, we propose a resampling strategy to Denoise using the user's Personal Loss distribution, named PLD, which aims to reduce the probability of noisy interactions being optimized. Specifically, during each optimization iteration, we create a candidate item pool for each user and resample the items from this pool based on the user's personal loss distribution, prioritizing normal interactions. Additionally, we conduct a theoretical analysis to validate PLD's effectiveness and suggest ways to further enhance its performance. Extensive experiments conducted on three datasets with varying noise ratios demonstrate PLD's efficacy and robustness.

Graph Neural Networks (GNNs) have achieved remarkable success in node classification tasks on individual graphs. However, existing GNNs trained within a specific domain (a.k.a., source domain) frequently exhibit unsatisfied performance when transferred to another domain (a.k.a., target domain), due to the domain gap. To tackle this issue, Few Shot Graph Domain Adaptation (FSGDA) is introduced to the node classification task, facilitating knowledge transfer from a fully labeled source graph to a target graph with minimal annotations for each class. An intuitive solution is directly training the GNN with labeled source and target samples together. Nevertheless, there are two issues in this procedure: (1) When the annotations on the target domain used for training are extremely sparse, the GNN performance may significantly be damaged by nodes with the source-domain bias not aligning with the target-domain distribution. (2) Apart from the biased nodes, the low-value nodes among the remaining nodes impede the GNN learning for the core nodes, like the limited target training nodes. To address the above issues, we propose a new method for FSGDA, named GraphInflu, whose core idea is to grasp the key takeaways from the source domain to facilitate the adaptation process. It contains two characteristic modules, including the Supportive Node Selector and the Soft Logic-Inspired Node Reweighting. The former aims to identify the most influential set of source nodes based on their contribution to improving performance on target nodes. The latter further focuses more on the core nodes in the selected influential set, which closely align with the target nodes especially those presenting challenging predictions. Extensive experiments validate the efficacy of GraphInflu by overcoming the current state-of-the-art methods. Our code is available at https://anonymous.4open.science/r/GraphInflu-E8E7.

We study sequential procurement auctions where the sellers are provided with a "best and final offer" (BAFO) strategy. This strategy allows each seller $i$ to effectively ``freeze'' their price while remaining active in the auction, and it signals to the buyer, as well as all other sellers, that seller $i$ would reject any price lower than that. This is in contrast to prior work, e.g., on descending auctions, where the options provided to each seller are to either accept a price reduction or reject it and drop out. As a result, the auctions that we consider induce different extensive form games and our goal is to study the subgame perfect equilibria of these games. We focus on settings involving multiple sellers who have full information regarding each other's cost (i.e., the minimum price that they can accept) and a single buyer (the auctioneer) who has no information regarding these costs. Our main result shows that the auctions enhanced with the BAFO strategy can guarantee efficiency in every subgame perfect equilibrium, even if the buyer's valuation function is an arbitrary monotone function. This is in contrast to prior work which required that the buyer's valuation satisfies restrictive properties, like gross substitutes, to achieve efficiency. We then also analyze the seller's cost in these subgame perfect equilibria and we show that it can vary significantly across different efficient outcomes, depending on the structure of the buyer's valuation function.

Augmenting large language models (LLMs) with external tools has emerged as a promising approach to extend their utility, enabling them to solve practical tasks. Previous methods manually parse tool documentation and create in-context demonstrations, transforming tools into structured formats for LLMs to use in their step-by-step reasoning. However, this manual process requires domain expertise and struggles to scale to large toolsets. % The LLMs show diminished performance when in-context examples are incomplete. Additionally, these methods rely heavily on ad-hoc inference technique or special tokens to integrate free-form LLM generation with tool-calling actions, limiting the LLM's flexibility in handling diverse tool specifications and integrating multiple tools. In this work, we propose AutoTools, a framework that enables LLMs to automate the tool-use workflow. Specifically, the LLM automatically transforms tool documentation into callable functions, verifying syntax and runtime correctness. Then, the LLM integrates these functions into executable programs to solve practical tasks, flexibly grounding tool-use actions into its reasoning processes. Extensive experiments on existing and newly collected, more challenging benchmarks illustrate the superiority of our framework. Inspired by these promising results, we further investigate how to improve the expertise of LLMs, especially open-source LLMs with fewer parameters, within AutoTools. Thus, we propose a method for AutoTools-learning, training the LLMs with three learning tasks on 34k instances of high-quality synthetic data, including documentation understanding, relevance learning and function programming. Fine-grained results validate the effectiveness of our overall training approach and each individual task. Our methods are an important step towards the use of LLMs for solving real-world tasks with external tools.

Web and mobile systems show constant distribution shifts due to the evolvement of services, users, and threats, severely degrading the performance of threat detection models trained on prior distributions. Fast model adaptation with minimal data from new distributions is essential for maintaining reliable security measures. A key challenge in this context is the lack of ground truth, which undermines the ability of existing solutions to align classes across shifted distributions. Moreover, the limited new data often fails to represent the underlying distribution, providing sparse and potentially noisy information for adaptation. In this paper, we propose REACT, a novel framework that adapts model weights using a few unlabeled data and contextual insights. We leverage the inherent data imbalance in threat detection and meta-train weights to generalize majority patterns across varying distributions, eliminating the reliance on labels for alignment. REACT decomposes a neural network into two complementary components: meta weights as a shared foundation of general knowledge, and residual adaptive weights as adjustments for specific shifts. To compensate for the limited availability of new data, REACT trains a hypernetwork to predict adaptive weights based on data and contextual information, enabling knowledge sharing across distributions. The meta weights and the hypernetwork are updated alternately, maximizing both generalization and adaptability. REACT is model-agnostic, applicable to various neural networks. We provide convergence analysis and conduct extensive experiments across multiple datasets and models. REACT improves AUROC by 14.85% over models without adaptation, outperforming the state-of-the-art.

Personalized algorithms can inadvertently expose users to discomforting recommendations, potentially triggering negative consequences. The subjectivity of discomfort and the black-box nature of these algorithms make it challenging to effectively identify and filter such content. To address this, we first conducted a formative study to understand users' practices and expectations regarding discomforting recommendation filtering. Then, we designed a Large Language Model (LLM)-based tool named DiscomfortFilter, which constructs an editable preference profile for a user and helps the user express filtering needs through conversation to mask discomforting preferences within the profile. Based on the edited profile, DiscomfortFilter facilitates the discomforting recommendations filtering in a plug-and-play manner, maintaining flexibility and transparency. The constructed preference profile improves LLM reasoning and simplifies user alignment, enabling a 3.8B open-source LLM to rival top commercial models in an offline proxy task. A one-week user study with 24 participants demonstrated the effectiveness of DiscomfortFilter, while also highlighting its potential impact on platform recommendation outcomes. We conclude by discussing the ongoing challenges, highlighting its relevance to broader research, assessing stakeholder impact, and outlining future research directions.