Explaining anomalies in spatiotemporal data requires more than identifying statistical deviations; it often demands coherent narratives that relate events to their spatial, temporal, and social context. In urban settings, where anomalies may reflect public events, behavioral shifts, or environmental factors, the lack of interpretable, contextual explanations limits the effectiveness of automated analysis. We present GroundSTory , an interactive framework that reframes spatiotemporal anomaly explanation as computational storytelling, integrating graph-based representations of urban space, coordinated visual analytics, and large language models (LLMs) to support the construction of grounded, human-readable narrative explanations. Anomalies act as narrative anchors, while localized spatial and temporal relations and contextual signals derived from the graph guide and constrain narrative generation. The LLM functions as a narrative assistant, proposing plausible, evidence-backed hypotheses that analysts can inspect, compare, and validate through visual interaction. We illustrate the approach through qualitative usage scenarios on real-world urban datasets from São Paulo and New York City, showing how interactive narratives surface meaningful links between detected anomalies and contextual factors such as holidays, mass events, and routine activity patterns. A prompt analysis further suggests that structured, graph-derived context tends to affect the stability and perceived quality of generated hypotheses. Overall, our results indicate that combining knowledge-driven representations, interactive visualization, and constrained prompting can support more coherent, plausible, and interpretable narratives for spatiotemporal anomaly analysis, contributing to ongoing research in interactive storytelling and narrative-centered AI systems.
Modeling user intent across heterogeneous domains remains a central challenge in cross-domain sequential recommendation (CDSR). Although cross-domain sequences naturally embed richer behavioral cues that traverse multiple domains, existing approaches often struggle to disentangle their distinctive effects from those of single-domain interactions. To address this limitation, this study proposes a novel user modeling framework that explicitly decomposes user preferences into three orthogonal components: domain-specific, domain-common, and cross-sequence-exclusive representations. The proposed method employs a triple-encoder architecture, where each encoder is specialized to capture one of these preference types. This is achieved through domain masking objectives and adversarial learning with a gradient reversal layer to promote balanced separation. In addition, representational alignment and mutual independence are jointly enforced to achieve a clean disentanglement among the components. To further enhance temporal adaptability, a dynamic gating mechanism is incorporated to modulate the relative influence of each preference type at every timestep, culminating in a unified session-level representation. Extensive experiments conducted on two large-scale real-world benchmarks confirm that the proposed approach not only enhances recommendation accuracy but also yields interpretable insights into the interplay between domain-specific and cross-domain preference signals.
The ancient art of Traditional Chinese Medicine (TCM) is distinguished by its extensive annals of utilizing nature's botanical emissaries to ameliorate various pathological conditions. TCM diagnosis and treatment exhibit a pronounced degree of personalization and an organically holistic approach, necessitating a comprehensive consideration of the patient's evolving physiological and symptomatic states over temporal domains. However, existing methods for TCM recommendation represent a flaw in accounting for the dynamic oscillations in patients’ conditions, instead restricting their explorations to potential correlative patterns between symptomatic presentations and prescribed remedies. In this paper, we propose a novel approach for optimizing TCM recommendations based on Sequential Condition Evolution and Interaction Knowledge Graphs SCEIKG , an innovation framework that conceptualizes the method as a sequential prescription-making problem by considering the inherent dynamism of patients’ conditions across multiple diagnoses. Furthermore, we incorporate an interaction knowledge graph to enhance the accuracy of recommendations by considering the intricate interplay between various herbs and patients’ states. Experimental results on the real-world dataset demonstrate that our approach outperforms existing TCM recommendation methods, achieving state-of-the-art performance. The source code of SCEIKG is available at https://github.com/jzephyrl/SCEIKG-TCM .
Causal effect estimation is fundamental to personalized decision-making and policy evaluation, with applications spanning healthcare, economics, and social sciences. However, observational data often suffer from selection bias and the absence of counterfactual outcomes, posing significant challenges to inference accuracy. While recent representation learning-based approaches have shown promise, they fail to fully exploit the rich self-supervised information and causal prior knowledge embedded in the data. To address these limitations, we propose S elf-supervised C ausal E ffects E stimation ( SCEE ), a novel framework that integrates causal priors with self-supervised learning to construct balanced and predictive representations for causal effects estimation. Experimental results on widely used real-world, semi-synthetic, and synthetic benchmarks demonstrate that SCEE consistently outperforms state-of-the-art methods. To further enhance its effectiveness, we investigate different contrastive sample selection strategies, maximizing the potential of contrastive learning in causal inference. Additionally, we analyze the impact of sample reweighting and show that SCEE inherently mitigates distributional discrepancies between treatment and control groups, eliminating the need for explicit reweighting mechanisms.
Traffic congestion remains a prevalent issue in urban areas, contributing to environmental pollution, increased fuel consumption, and delays in emergency services. Addressing this challenge is paramount, with traffic flow prediction emerging as a pivotal technology within Intelligent Transportation Systems (ITS) to mitigate congestion and conserve time and energy. Extensive efforts have been directed towards developing predictive models, categorized into parametric, non-parametric, and hybrid approaches. While existing literature has extensively explored parametric and non-parametric models, this survey paper focuses on recent advancements in traffic flow prediction models, examining their merits and drawbacks. A comprehensive taxonomy of hybrid models utilized in traffic prediction is also provided. Additionally, the role of contextual features in enhancing traffic flow prediction is explored. Further, popular datasets employed for training traffic flow prediction models are discussed, accompanied by a comparative analysis of leading models’ performance using these datasets. The survey explores various performance metrics employed to evaluate prediction model efficacy. Furthermore, it identifies research gaps and challenges and outlines future research directions, aiming to contribute to the ongoing discourse on traffic flow prediction and inform future endeavors in this domain.
