Ya Su, Hu Zhang, Dan Qiao, YuJie Wang, Yunxiao Zhao, Yue Fan, Shike Li, Ru Li, Hongye Tan. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Event Causality Identification (ECI) aims to identify fine-grained causal relationships between events in an unstructured text. Existing ECI methods primarily rely on knowledge enhanced and graph-based reasoning approaches, but they often overlook the dependencies between similar events. Additionally, the connection between unstructured text and structured knowledge is relatively weak. Therefore, this paper proposes an ECI method enhanced by LLM Knowledge and Concept-Level Event Relations (LKCER). Specifically, LKCER constructs a conceptual-level heterogeneous event graph by leveraging the local contextual information of related event mentions, generating a more comprehensive global semantic representation of event concepts. At the same time, the knowledge generated by COMET is filtered and enriched using LLM, strengthening the associations between event pairs and knowledge. Finally, the joint event conceptual representation and knowledge-enhanced event representation are used to uncover potential causal relationships between events. The experimental results show that our method outperforms previous state-of-the-art methods on both benchmarks, EventStoryLine and Causal-TimeBank.

Event Causal Identification (ECI) aims to identify fine-grained causal relationships between events from unstructured text. Contrastive learning has shown promise in enhancing ECI by optimizing representation distances between positive and negative samples. However, existing methods often rely on rule-based or random sampling strategies, which may introduce spurious causal positives. Moreover, static negative samples often fail to approximate actual decision boundaries, thus limiting discriminative performance. Therefore, we propose an ECI method enhanced by Dynamic Energy-based Contrastive Learning with multi-stage knowledge Verification (DECLV). Specifically, we integrate multi-source knowledge validation and LLM-driven causal inference to construct a multi-stage knowledge validation mechanism, which generates high-quality contrastive samples and effectively suppresses spurious causal disturbances. Meanwhile, we introduce the Stochastic Gradient Langevin Dynamics (SGLD) method to dynamically generate adversarial negative samples, and employ an energy-based function to model the causal boundary between positive and negative samples. The experimental results show that our method outperforms previous state-of-the-art methods on both benchmarks, EventStoryLine and Causal-TimeBank.

Network-based intrusion detection system (NIDS) monitors network traffic for malicious activities, forming the frontline defense against increasing attacks over information infrastructures. Although promising, our quantitative analysis shows that existing methods perform inconsistently in declaring various unknown attacks (e.g., 9% and 35% F1 respectively for two distinct unknown threats for an SVM-based method) or detecting diverse known attacks (e.g., 31% F1 for the Backdoor and 93% F1 for DDoS for a GCN-based state-of-the-art method), and reveals that the underlying cause is entangled distributions of flow features. This motivates us to propose 3D-IDS, a novel method that aims to tackle the above issues through two-step feature disentanglements and a dynamic graph diffusion scheme. Specifically, we first disentangle traffic features by a non-parameterized optimization based on mutual information, automatically differentiating tens and hundreds of complex features of various attacks. Such differentiated features will be fed into a memory model to generate representations, which are further disentangled to highlight the attack-specific features. Finally, we use a novel graph diffusion method that dynamically fuses the network topology for spatial-temporal aggregation in evolving data streams. By doing so, we can effectively identify various attacks in encrypted traffics, including unknown threats and known ones that are not easily detected. Experiments show the superiority of our 3D-IDS. We also demonstrate that our two-step feature disentanglements benefit the explainability of NIDS.

BACKGROUND: Stable Isotope Resolved Metabolomics (SIRM) is a new biological approach that uses stable isotope tracers such as uniformly [Formula: see text]-enriched glucose ([Formula: see text]-Glc) to trace metabolic pathways or networks at the atomic level in complex biological systems. Non-steady-state kinetic modeling based on SIRM data uses sets of simultaneous ordinary differential equations (ODEs) to quantitatively characterize the dynamic behavior of metabolic networks. It has been increasingly used to understand the regulation of normal metabolism and dysregulation in the development of diseases. However, fitting a kinetic model is challenging because there are usually multiple sets of parameter values that fit the data equally well, especially for large-scale kinetic models. In addition, there is a lack of statistically rigorous methods to compare kinetic model parameters between different experimental groups. RESULTS: We propose a new Bayesian statistical framework to enhance parameter estimation and hypothesis testing for non-steady-state kinetic modeling of SIRM data. For estimating kinetic model parameters, we leverage the prior distribution not only to allow incorporation of experts' knowledge but also to provide robust parameter estimation. We also introduce a shrinkage approach for borrowing information across the ensemble of metabolites to stably estimate the variance of an individual isotopomer. In addition, we use a component-wise adaptive Metropolis algorithm with delayed rejection to perform efficient Monte Carlo sampling of the posterior distribution over high-dimensional parameter space. For comparing kinetic model parameters between experimental groups, we propose a new reparameterization method that converts the complex hypothesis testing problem into a more tractable parameter estimation problem. We also propose an inference procedure based on credible interval and credible value. Our method is freely available for academic use at https://github.com/xuzhang0131/MCMCFlux . CONCLUSIONS: Our new Bayesian framework provides robust estimation of kinetic model parameters and enables rigorous comparison of model parameters between experimental groups. Simulation studies and application to a lung cancer study demonstrate that our framework performs well for non-steady-state kinetic modeling of SIRM data.

The Corona Virus Disease 2019 (COVID-19) epidemic is a sudden public health crisis, known as an "International Emergency of Public Health Event". This study uses the bottom-up characteristics of multi-agents to construct multi-agent simulation models for COVID-19 prevention and control. The development trend of the epidemic situation under the condition that the government adopts different prevention and control measures is studied, and on this basis, the influence of temperature on the spread of the virus is discussed. The simulation results show that the multi-agent modeling method can effectively capture the emergence of complex systems. The evaluation of the effects of single measures and multiple interventions will help determine key prevention and control strategies and provide important experience and scientific basis for future epidemic prevention and control.

System instance clustering is crucial for large-scale Web services because it can significantly reduce the training overhead of anomaly detection methods. However, the vast number of system instances with massive time points, redundant metrics, and noise bring significant challenges. We propose OmniCluster to accurately and efficiently cluster system instances for large-scale Web services. It combines a one-dimensional convolutional autoencoder (1D-CAE), which extracts the main features of system instances, with a simple, novel, yet effective three-step feature selection strategy. We evaluated OmniCluster using real-world data collected from a top-tier content service provider providing services for one billion+ monthly active users (MAU), proving that OmniCluster achieves high accuracy (NMI=0.9160) and reduces the training overhead of five anomaly detection models by 95.01% on average.

Today's large datacenters house a massive number of machines, each of which is being closely monitored with multivariate time series (e.g., CPU idle, memory utilization) to ensure service quality. Detecting outlier machine instances with multivariate time series is crucial for service management. However, it is a challenging task due to the multiple classes and various shapes, high dimensionality, and lack of labels of multivariate time series. In this article, we propose DOMI, a novel unsupervised model that combines Gaussian mixture VAE with 1D-CNN, todetectoutliermachineinstances. Its core idea is to capture the normal patterns of machine instances by learning their latent representations that consider the shape characteristics, reconstruct input data by the learned representations, and apply reconstruction probabilities to determine outliers. Moreover, DOMI interprets the detected outlier instance based on the reconstruction probability changes of univariate time series. Extensive experiments have been conducted on the dataset collected from 1821 machines with a 1.5-month-period, which are deployed in ByteDance, a top global content service provider. DOMI achieves the best F1-Score of 0.94 and AUC score of 0.99, significantly outperforming the best performing baseline method by 0.08 and 0.03, respectively. Moreover, its interpretation accuracy is up to 0.93.

Anomaly detection is indispensable in modern IT infrastructure management. However, the dimension explosion problem of the monitoring data (large-scale machines, many key performance indicators, and frequent monitoring queries) causes a scalability issue to the existing algorithms. We propose a coarse-to-fine model transfer based framework CTF to achieve a scalable and accurate data-center-scale anomaly detection. CTF pre-trains a coarse-grained model, uses the model to extract and compress per-machine features to a distribution, clusters machines according to the distribution, and conducts model transfer to fine-tune per-cluster models for high accuracy. The framework takes advantage of clustering on the per-machine latent representation distribution, reusing the pre-trained model, and partial-layer model fine-tuning to boost the whole training efficiency. We also justify design choices such as the clustering algorithm and distance algorithm to achieve the best accuracy. We prototype CTF and experiment on production data to show its scalability and accuracy. We also release a labeling tool for multivariate time series and a labeled dataset to the research community.

To ensure the reliability of software services, operators collect and monitor a large number of KPI (Key Performance Indicator) streams constantly. KPI anomaly detection is vitally important for software service management. However, none of supervised learning methods, semi-supervised learning methods, transfer learning methods, or unsupervised learning methods achieve accurate anomaly detection for the large-scale, diverse, dynamically changing KPI streams with little labeling effort. In this paper, we propose PUAD, a PU learning-based method, to achieve accurate KPI anomaly detection requiring a few partial labels. It integrates clustering, PU learning, and semi-supervised learning to minimize labeling effort and improve anomaly detection accuracy simultaneously. Additionally, we propose a novel active learning method that selects the samples most likely to be positive in each iteration to avoid false alarms. We apply 208 real-world KPI streams collected from a large-scale software service provider to evaluate the performance of PUAD, demonstrating that it achieves a close F1-score to supervised learning methods with much fewer manual labels, and greatly outperforms semi-supervised learning methods, transfer learning methods, and unsupervised learning methods.

Anomaly detection is a crucial task for monitoring various status (i.e., metrics) of entities (e.g., manufacturing systems and Internet services), which are often characterized by multivariate time series (MTS). In practice, it's important to precisely detect the anomalies, as well as to interpret the detected anomalies through localizing a group of most anomalous metrics, to further assist the failure troubleshooting. In this paper, we propose InterFusion, an unsupervised method that simultaneously models the inter-metric and temporal dependency for MTS. Its core idea is to model the normal patterns inside MTS data through hierarchical Variational AutoEncoder with two stochastic latent variables, each of which learns low-dimensional inter-metric or temporal embeddings. Furthermore, we propose an MCMC-based method to obtain reasonable embeddings and reconstructions at anomalous parts for MTS anomaly interpretation. Our evaluation experiments are conducted on four real-world datasets from different industrial domains (three existing and one newly published dataset collected through our pilot deployment of InterFusion). InterFusion achieves an average anomaly detection F1-Score higher than 0.94 and anomaly interpretation performance of 0.87, significantly outperforming recent state-of-the-art MTS anomaly detection methods.

Internet-based service companies monitor a large number of KPIs (Key Performance Indicators) to ensure their service quality and reliability. Correlating KPIs by fluctuations reveals interactions between KPIs under anomalous situations and can be extremely useful for service troubleshooting. However, such a KPI flux-correlation has been little studied so far in the domain of Internet service operations management. A major challenge is how to automatically and accurately separate fluctuations from normal variations in KPIs with different structural characteristics (such as seasonal, trend and stationary) for a large number of KPIs. In this paper, we propose CoFlux, an unsupervised approach, to automatically (without manual selection of algorithm fitting and parameter tuning) determine whether two KPIs are correlated by fluctuations, in what temporal order they fluctuate, and whether they fluctuate in the same direction. CoFlux's robust feature engineering and robust correlation score computation enable it to work well against the diverse KPI characteristics. Our extensive experiments have demonstrated that CoFlux achieves the best F1-Scores of 0.84 (0.90), 0.92 (0.95), 0.95 (0.99), in answering these three questions, in the two real datasets from a top global Internet company, respectively. Moreover, we showed that CoFlux is effective in assisting service troubleshooting through the applications of alert compression, recommending Top N causes, and constructing fluctuation propagation chains.

Industry devices (i.e., entities) such as server machines, spacecrafts, engines, etc., are typically monitored with multivariate time series, whose anomaly detection is critical for an entity's service quality management. However, due to the complex temporal dependence and stochasticity of multivariate time series, their anomaly detection remains a big challenge. This paper proposes OmniAnomaly, a stochastic recurrent neural network for multivariate time series anomaly detection that works well robustly for various devices. Its core idea is to capture the normal patterns of multivariate time series by learning their robust representations with key techniques such as stochastic variable connection and planar normalizing flow, reconstruct input data by the representations, and use the reconstruction probabilities to determine anomalies. Moreover, for a detected entity anomaly, OmniAnomaly can provide interpretations based on the reconstruction probabilities of its constituent univariate time series. The evaluation experiments are conducted on two public datasets from aerospace and a new server machine dataset (collected and released by us) from an Internet company. OmniAnomaly achieves an overall F1-Score of 0.86 in three real-world datasets, signicantly outperforming the best performing baseline method by 0.09. The interpretation accuracy for OmniAnomaly is up to 0.89.

Face recognition technique has obtained great progress and excellent results on public data sets. However, traditional algorithms suffer from various changes such as illumination, expression, and misalignment in practical applications. To solve these problems, this study proposes a novel face recognition algorithm simultaneously resolves these challenges. The key idea is reducing the influence of illumination and expression through the aligning procedure. As a result, illumination, expression, and misalignment can be greatly ignored in the recognition procedure. The contributions of this study are two folds. (i) The construction of the shape constrained illumination pattern (SCIP), which models the illumination variation with robustness to expression change. (ii) SCIP‐based face recognition algorithm which can deal with illumination, expression, and image misalignment simultaneously. Systematic evaluations conducted on public databases demonstrate that the proposed algorithm is robust to illumination, expression, and misalignment with better performance than state‐of‐the‐art algorithms.

Video text extraction plays an important role for multimedia understanding and retrieval. Most previous research efforts are conducted within individual frames. A few of recent methods, which pay attention to text tracking using multiple frames, however, do not effectively mine the relations among text detection, tracking and recognition. In this paper, we propose a generic Bayesian-based framework of Tracking based Text Detection And Recognition (T DAR) from web videos for embedded captions, which is composed of three major components, i.e., text tracking, tracking based text detection, and tracking based text recognition. In this unified framework, text tracking is first conducted by tracking-by-detection. Tracking trajectories are then revised and refined with detection or recognition results. Text detection or recognition is finally improved with multi-frame integration. Moreover, a challenging video text (embedded caption text) database (USTB-VidTEXT) is constructed and publicly available. A variety of experiments on this dataset verify that our proposed approach largely improves the performance of text detection and recognition from web videos.

Previous studies have shown the WiFi, as the dominant last hop access to Internet, has become the weakest link in the round-trip network delay. Therefore it is critical to understand and minimize the WiFi interference in order to reduce the WiFi hop delay. For the first time in the literature, this paper defines an intuitive and accurate metric to quantify the impact of interference on each actual packet. For each packet traveling through the access point, it measures the percentage of MAC layer delay wasted due to neighbor APs' interference. This metric is defined based on a packet's various (measured or inferred) timestamps and can be measured with a small kernel modification on a commodity AP with little overhead. Our 29-AP two- month measurement results in the wild show that this metric is a strong indicator of interference's impact on WiFi hop delay. Using this metric as input, distributed channel selection on individual APs reduces the median WiFi hop delay by up to 5X. Collaborative optimization on multiple APs reduces the overall WiFi hop delay by 5X compared to the default channel.

Studies have discovered that face recognition will benefit from age information. However, since the age estimation is unstable in practice, it is still an open question how to improve face recognition with help of automatic age estimation techniques. This paper presents to improve the performance of face recognition by automatic age estimation. The main contribution is a new age-variational face recognition algorithm based on Bayesian framework (FRAB). By introducing the age estimation result as a prior, the recognition problem is divided into several age-specific sub-problems. As a result, the proposed algorithm leads to two algorithms according to how the age is given. The first one is FRAB-AE, which introduces age estimation result as the age prior. The second one is FRAB-GT, which considers that the ground truth of age information is given. Experimental results are conducted on FG-NET and Morph datasets to evaluate the performance of the proposed framework. It shows that the proposed algorithms is able to make use of age priors to improve the face recognition.

Low-dimensional structures embedded in high-dimensional data space can be extracted by spectral analysis and manifold learning. Standard approaches to manifold learning are usually based on the assumption that there is a dominant low-dimensional manifold, while other variations are considered with minor priority. We instead consider the scenario that a pair of distinct manifolds intertwined in the same high-dimensional space, which can be decomposed for analysis. The core of this new method is a novel submanifold decomposition (SMD) algorithm. This paper has three contributions: 1) a submanifold framework is proposed to model the high-dimensional dataset, which is dominated by more than one factor; 2) a nonlinear manifold decomposition method, SMD, is presented to extract two intertwined manifolds from a dataset in a discriminative manner; and 3) in order to solve the out-of-sample problem of nonlinear SMD, a linear extension of SMD is developed, which is effective to extract two linear submanifolds. We demonstrate that comparing with the existing manifold learning methods that only extract one dominant manifold, the proposed SMD and its linear extension are capable of extracting a pair of submanifolds discriminatively and effectively. Moreover, the two extracted manifolds can complement each other to enhance the representation performance. Extensive experiments on both artificial data and real data demonstrate that the proposed method outperforms the state-of-the-art manifold learning algorithms in visual recognition tasks.

Canonical correlation has been prevalent for multiset-based pairwise subspace analysis. As an extension, discriminant canonical correlations (DCCs) have been developed for classification purpose by learning a global subspace based on Fisher discriminant modeling of pairwise subspaces. However, the discriminative power of DCCs is not optimal as it only measures the "local" canonical correlations within subspace pairs, which lacks the "global" measurement among all the subspaces. In this paper, we propose a multiset discriminant canonical correlation method, i.e., multiple principal angle (MPA). It jointly considers both "local" and "global" canonical correlations by iteratively learning multiple subspaces (one for each set) as well as a global discriminative subspace, on which the angle among multiple subspaces of the same class is minimized while that of different classes is maximized. The proposed computational solution is guaranteed to be convergent with much faster converging speed than DCC. Extensive experiments on pattern recognition applications demonstrate the superior performance of MPA compared to existing subspace learning methods.

Conventional regression methods, such as multivariate linear regression (MLR) and its extension principal component regression (PCR), deal well with the situations that the data are of the form of low-dimensional vector. When the dimension grows higher, it leads to the under sample problem (USP): the dimensionality of the feature space is much higher than the number of training samples. However, little attention has been paid to such a problem. This paper first adopts an in-depth investigation to the USP in PCR, which answers three questions: 1) Why is USP produced? 2) What is the condition for USP, and 3) How is the influence of USP on regression. With the help of the above analysis, the principal components selection problem of PCR is presented. Subsequently, to address the problem of PCR, a multivariate multilinear regression (MMR) model is proposed which gives a substitutive solution to MLR, under the condition of multilinear objects. The basic idea of MMR is to transfer the multilinear structure of objects into the regression coefficients as a constraint. As a result, the regression problem is reduced to find two low-dimensional coefficients so that the principal components selection problem is avoided. Moreover, the sample size needed for solving MMR is greatly reduced so that USP is alleviated. As there is no closed-form solution for MMR, an alternative projection procedure is designed to obtain the regression matrices. For the sake of completeness, the analysis of computational cost and the proof of convergence are studied subsequently. Furthermore, MMR is applied to model the fitting procedure in the active appearance model (AAM). Experiments are conducted on both the carefully designed synthesizing data set and AAM fitting databases verified the theoretical analysis.

Due to the temporal property of age progression, face images with agingingg features display some sequential patterns with low-dimensional distributions, which can be effectively extracted by subspace learning algorithms. The patterns extracted by traditional subspace learning methods are mostly restricted to a certain database. As a result, the performance cannot be generalized when applying these patterns to cross databases with different multi-mode variations (e.g. gender, identity, and imaging conditions.) This problem has yet not been given much attention before. In this paper, the cross-database age estimation problem is solved by a transfer learning framework. The proposed framework transfers the knowledge gained from training samples to the target data and improves the performance in cross-database scenarios. Experimental results for age estimation tasks on different datasets demonstrate the effectiveness and robustness of our proposed framework.

Active appearance model (AAM) is a powerful generative method for modeling deformable objects. The model decouples the shape and the texture variations of objects, which is followed by an efficient gradient-based model fitting method. Due to the flexible and simple framework, AAM has been widely applied in the fields of computer vision. However, difficulties are met when it is applied to various practical issues, which lead to a lot of prominent improvements to the model. Nevertheless, these difficulties and improvements have not been studied systematically. This motivates us to review the recent advances of AAM. This paper focuses on the improvements in the literature in turns of the problems suffered by AAM in practical applications. Therefore, these algorithms are summarized from three aspects, i.e., efficiency, discrimination, and robustness. Additionally, some applications and implementations of AAM are also enumerated. The main purpose of this paper is to serve as a guide for further research.

Active appearance model (AAM) has been widely used for modeling the shape and the texture of deformable objects and matching new ones effectively. The traditional AAM consists of two parts, shape model and texture model. In the texture model, for the sake of simplicity, the image intensity is usually employed to represent the texture information. However, the intensity is easy to be interfered by the external environment change, e.g. illumination variations, which results in an unsatisfied model fitting. To this purpose, we present a new texture representation in AAM, which combines Gabor wavelet and local binary patterns (LBP) operator. On the one hand, Gabor wavelet can encode multi-scale and multi-direction information of an image. On the other hand, LBP is able to efficiently encode local information and compress the redundancy in the Gabor filtered images. Since the new texture representation can express an object more sophisticatedly, it will improve the accuracy of the model fitting. The experimental results on various datasets demonstrate the effectiveness of the proposed texture representation, which results in a more accurate and reliable matching between the model and new images.

Active Appearance Models (AAMs) are generative models which can describe deformable objects. However, the texture in basic AAMs is represented using intensity values. Despite its simplicity, this representation does not contain enough information for image matching. In this paper, we firstly propose to utilize Gabor filters to represent the image texture. The benefit of Gabor-based representation is that it can express local structures of an image. As a result, this representation can lead to more accurate matching when condition changes. Given the problem of the excessive storage and computational complexity of the Gabor, three different Gabor-based image representations are used in AAMs: (1) GaborD is the sum of Gabor filter responses over directions, (2) GaborS is the sum of Gabor filter responses over scales, and (3) GaborSD is the sum of Gabor filter responses over scales and directions. Through a large number of experiments, we show that the proposed Gabor representations lead to more accurate and reliable matching between model and images.