Deep learning’s performance has been extensively recognized recently. Graph neural networks (GNNs... more Deep learning’s performance has been extensively recognized recently. Graph neural networks (GNNs) are designed to deal with graph-structural data that classical deep learning does not easily manage. Since most GNNs were created using distinct theories, direct comparisons are impossible. Prior research has primarily concentrated on categorizing existing models, with little attention paid to their intrinsic connections. The purpose of this study is to establish a unified framework that integrates GNNs based on spectral graph and approximation theory. The framework incorporates a strong integration between spatial- and spectral-based GNNs while tightly associating approaches that exist within each respective domain.
Proceedings of the AAAI Conference on Artificial Intelligence
Data-driven related solutions are dominating various scientific fields with the assistance of mac... more Data-driven related solutions are dominating various scientific fields with the assistance of machine learning and data analytics. Finding effective solutions has long been discussed in the area of machine learning. The recent decade has witnessed the promising performance of the Physics-Informed Neural Networks (PINN) in bridging the gap between real-world scientific problems and machine learning models. In this paper, we explore the behavior of PINN in a particular range of different diffusion coefficients under specific boundary conditions. In addition, different initial conditions of partial differential equations are solved by applying the proposed PINN. Our paper illustrates how the effectiveness of the PINN can change under various scenarios. As a result, we demonstrate a better insight into the behaviors of the PINN and how to make the proposed method more robust while encountering different scientific and engineering problems.
Proceedings of the AAAI Conference on Artificial Intelligence
Diverse efforts to combat the COVID-19 pandemic have continued throughout the past two years. Gov... more Diverse efforts to combat the COVID-19 pandemic have continued throughout the past two years. Governments have announced plans for unprecedentedly rapid vaccine development, quarantine measures, and economic revitalization. They contribute to a more effective pandemic response by determining the precise opinions of individuals regarding these mitigation measures. In this paper, we propose a deep learning-based topic monitoring and storyline extraction system for COVID-19 that is capable of analyzing public sentiment and pandemic trends. The proposed method is able to retrieve Twitter data related to COVID-19 and conduct spatiotemporal analysis. Furthermore, a deep learning component of the system provides monitoring and modeling capabilities for topics based on advanced natural language processing models. A variety of visualization methods are applied to the project to show the distribution of each topic. Our proposed system accurately reflects how public reactions change over time ...
The prompt estimation of traffic incident impacts can guide commuters in their trip planning and ... more The prompt estimation of traffic incident impacts can guide commuters in their trip planning and improve the resilience of transportation agencies' decision-making on resilience. However, it is more challenging than node-level and graphlevel forecasting tasks, as it requires extracting the anomaly subgraph or sub-time-series from dynamic graphs. In this paper, we propose DG-Trans, a novel traffic incident impact prediction framework, to foresee the impact of traffic incidents through dynamic graph learning. The proposed framework contains a dual-level spatial transformer and an importance-score-based temporal transformer, and the performance of this framework is justified by two newly constructed benchmark datasets. The duallevel spatial transformer removes unnecessary edges between nodes to isolate the affected subgraph from the other nodes. Meanwhile, the importance-score-based temporal transformer identifies abnormal changes in node features, causing the predictions to rely more on measurement changes after the incident occurs. Therefore, DG-Trans is equipped with dual abilities that extract spatiotemporal dependency and identify anomaly nodes affected by incidents while removing noise introduced by benign nodes. Extensive experiments on real-world datasets verify that DG-Trans outperforms the existing state-of-the-art methods, especially in extracting spatiotemporal dependency patterns and predicting traffic accident impacts. It offers promising potential for traffic incident management systems. Index Terms-spatiotemporal data mining, intelligent transportation systems, incident impact forecasting, transformer, anomaly detection.
Proceedings of the AAAI Conference on Artificial Intelligence
Forecasting citations of scientific patents and publications is a crucial task for understanding ... more Forecasting citations of scientific patents and publications is a crucial task for understanding the evolution and development of technological domains and for foresight into emerging technologies. By construing citations as a time series, the task can be cast into the domain of temporal point processes. Most existing work on forecasting with temporal point processes, both conventional and neural network-based, only performs single-step forecasting. In citation forecasting, however, the more salient goal is n-step forecasting: predicting the arrival time and the technology class of the next n citations. In this paper, we propose Dynamic Multi-Context Attention Networks (DMA-Nets), a novel deep learning sequence-to-sequence (Seq2Seq) model with a novel hierarchical dynamic attention mechanism for long-term citation forecasting. Extensive experiments on two real-world datasets demonstrate that the proposed model learns better representations of conditional dependencies over historical...
Proceedings of the AAAI Conference on Artificial Intelligence
Predicting and quantifying the impact of traffic accidents is necessary and critical to Intellige... more Predicting and quantifying the impact of traffic accidents is necessary and critical to Intelligent Transport Systems (ITS). As a state-of-the-art technique in graph learning, current graph neural networks heavily rely on graph Fourier transform, assuming homophily among the neighborhood. However, the homophily assumption makes it challenging to characterize abrupt signals such as traffic accidents. Our paper proposes an abrupt graph wavelet network (AGWN) to model traffic accidents and predict their time durations using only one single snapshot.
Proceedings of the AAAI Conference on Artificial Intelligence
Chinese characters have semantic-rich compositional information in radical form. While almost all... more Chinese characters have semantic-rich compositional information in radical form. While almost all previous research has applied CNNs to extract this compositional information, our work utilizes deep graph learning on a compact, graph-based representation of Chinese characters. This allows us to exploit temporal information within the strict stroke order used in writing characters. Our results show that our stroke-based model has potential for helping large-scale language models on some Chinese natural language understanding tasks. In particular, we demonstrate that our graph model produces more interpretable embeddings shown through word subtraction analogies and character embedding visualizations.
Proceedings of the AAAI Conference on Artificial Intelligence
Influence blocking maximization (IBM) is crucial in many critical real-world problems such as rum... more Influence blocking maximization (IBM) is crucial in many critical real-world problems such as rumors prevention and epidemic containment. The existing work suffers from: (1) concentrating on uniform costs at the individual level, (2) mostly utilizing greedy approaches to approximate optimization, (3) lacking a proper graph representation for influence estimates. To address these issues, this research introduces a neural network model dubbed Neural Influence Blocking (\algo) for improved approximation and enhanced influence blocking effectiveness. The code is available at https://github.com/oates9895/NIB.
2020 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM)
Open and accessible public utilities such as mass public transit systems are some of the vexing v... more Open and accessible public utilities such as mass public transit systems are some of the vexing venues that are vulnerable to several criminal acts due to the large volumes of commuters. Existing forms of threat or event detection for the rail-based transit systems are either not working in real-time or do not provide complete coverage. In this paper, we present RISECURE 1 , an open-source system, that uses real-time social media mining to aid in the early detection of such possible events within a rail-based/metro system. The system leverages dynamic query expansion to keep track of any new emerging information about any particular incident. The Real Time Incident panel of the proposed system provides a comprehensible representation of the evolution of threatening transit events, which are further shown in the storyline modal for each respective station. The alert notification module of the system is capable of monitoring threats to the rail-based/metro systems in real-time. We demonstrate the system by including case studies involving incidents occurring within the Washington DC Metropolitan Area Transit Authority (WMATA) metro system to justify the effectiveness of our approach.
Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence
Modeling and forecasting forward citations to a patent is a central task for the discovery of eme... more Modeling and forecasting forward citations to a patent is a central task for the discovery of emerging technologies and for measuring the pulse of inventive progress. Conventional methods for forecasting these forward citations cast the problem as analysis of temporal point processes which rely on the conditional intensity of previously received citations. Recent approaches model the conditional intensity as a chain of recurrent neural networks to capture memory dependency in hopes of reducing the restrictions of the parametric form of the intensity function. For the problem of patent citations, we observe that forecasting a patent's chain of citations benefits from not only the patent's history itself but also from the historical citations of assignees and inventors associated with that patent. In this paper, we propose a sequence-to-sequence model which employs an attention-of-attention mechanism to capture the dependencies of these multiple time sequences. Furthermore, th...
is paper presents SAFEBIKE, a novel route recommendation system for bike-sharing service that uti... more is paper presents SAFEBIKE, a novel route recommendation system for bike-sharing service that utilizes station information to infer the number of available bikes in dock and recommend bike routes according to multiple factors such as distance and safety level. e system consists of a station level availability predictor that predicts bikes and docks amount at each station, and an efcient route recommendation service that considers safety and bike/dock availability factors. It targets users who are concerned about route safeness and station availability. We demonstrate the system by utilizing Citi Bike station availability and New York City crime data of Manha an to show the e ectiveness of our approach. Integrated with real-time station availability and historical crime data resources, our proposed system can e ectively recommend an optimal bike route and improve the travel experience of bike users.
2018 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM), 2018
With the rapid growth in urban transit networks in recent years, detecting service disruptions in... more With the rapid growth in urban transit networks in recent years, detecting service disruptions in a timely manner is a problem of increased interest to service providers. Transit agencies are seeking to move beyond traditional customer questionnaires and manual service inspections to leveraging open source indicators like social media for deteting emerging transit events. In this paper, we leverage Twitter data for early detection of metro service disruptions. Inspired by the multi-task learning framework, we propose the Metro Disruption Detection Model, which captures the semantic similarity between transit lines in Twitter space. We propose novel constraints on feature semantic similarity exploiting prior knowledge about the spatial connectivity and shared tracks of the metro network. An algorithm based on the alternating direction method of multipliers (ADMM) framework is developed to solve the proposed model. We run extensive experiments and comparisons to other models with real world Twitter data and transit disruption records from the Washington Metropolitan Area Transit Authority (WMATA) to justify the efficacy of our model.
A Hierarchical Attention Graph Convolutional Network for Traffic Incident Impact Forecasting
2021 IEEE International Conference on Big Data (Big Data), 2021
Predicting the impact of traffic i ncidents b ased on traffic s ensor d ata i s a n e ssential r ... more Predicting the impact of traffic i ncidents b ased on traffic s ensor d ata i s a n e ssential r esearch t opic i n t he fi eld of Intelligent Transportation Systems (ITS). Tackling the problem of estimating the durations of incidents from their early stages is a challenge due to the variable nature of such incidents and the complex structure of modern road networks. Existing studies on forecasting the incident duration from sensor data are mostly incapable of modeling 1) the spatiotemporal correlations of traffic s ensors a nd a rterial r oads a nd 2 ) t he hierarchical topology of the traffic sensor and road networks. In this paper, we propose the Hierarchical Attention-based Spatiotemporal Graph Convolutional Network model (HastGCN) to solve the incident duration forecasting problem by formulating the spatiotemporal correlation and traffic p atterns o n b oth t he s ensor l evel and the road level in their natural hierarchical manner. At the sensor level, we propose a spatiotemporal attention mechanism followed by graph convolutions to model the local correlations and patterns between traffic s ensors o n t he s ame a rterial road. At the road level, a connectivity-aware attention mechanism is designed to learn the global spatial relatedness between each arterial road. Traffic-condition a ware g raph c onvolutions are then applied to understand the target incident representation for the incident duration forecasting.
Proceedings of the 23rd SIGSPATIAL International Conference on Advances in Geographic Information Systems, 2015
Traditional statistical analysis on speed, volume, and occupancy has dominated the field of Arter... more Traditional statistical analysis on speed, volume, and occupancy has dominated the field of Arterial Incident Management Study (AIMS). However, few previous works have focused on investigating into the causality of the incidents. In this paper, we present ButterFly, a social media based arterial incident detection and analysis system. The proposed system is dedicated to identify the traffic incident from a novel perspective and discover causalities between traffic incidents. The main functionalities of the proposed system include: 1) Traffic incident detection based on userinput social media contents, 2) Transportation incidents storyline generation, and 3) Traffic incidents causalities analysis and visualization. We demonstrate the system by considering the Washington DC area as our experimental environment. ButterFly is targeted to provide effective and convenient real-time and historical traffic incidents analysis interfaces for transportation management agencies and academies. Our proposed system, integrated with multiple social media resources, can greatly broaden the visions for traffic incidents analysis.
Deep learning’s performance has been extensively recognized recently. Graph neural networks (GNNs... more Deep learning’s performance has been extensively recognized recently. Graph neural networks (GNNs) are designed to deal with graph-structural data that classical deep learning does not easily manage. Since most GNNs were created using distinct theories, direct comparisons are impossible. Prior research has primarily concentrated on categorizing existing models, with little attention paid to their intrinsic connections. The purpose of this study is to establish a unified framework that integrates GNNs based on spectral graph and approximation theory. The framework incorporates a strong integration between spatial- and spectral-based GNNs while tightly associating approaches that exist within each respective domain.
Proceedings of the AAAI Conference on Artificial Intelligence
Data-driven related solutions are dominating various scientific fields with the assistance of mac... more Data-driven related solutions are dominating various scientific fields with the assistance of machine learning and data analytics. Finding effective solutions has long been discussed in the area of machine learning. The recent decade has witnessed the promising performance of the Physics-Informed Neural Networks (PINN) in bridging the gap between real-world scientific problems and machine learning models. In this paper, we explore the behavior of PINN in a particular range of different diffusion coefficients under specific boundary conditions. In addition, different initial conditions of partial differential equations are solved by applying the proposed PINN. Our paper illustrates how the effectiveness of the PINN can change under various scenarios. As a result, we demonstrate a better insight into the behaviors of the PINN and how to make the proposed method more robust while encountering different scientific and engineering problems.
Proceedings of the AAAI Conference on Artificial Intelligence
Diverse efforts to combat the COVID-19 pandemic have continued throughout the past two years. Gov... more Diverse efforts to combat the COVID-19 pandemic have continued throughout the past two years. Governments have announced plans for unprecedentedly rapid vaccine development, quarantine measures, and economic revitalization. They contribute to a more effective pandemic response by determining the precise opinions of individuals regarding these mitigation measures. In this paper, we propose a deep learning-based topic monitoring and storyline extraction system for COVID-19 that is capable of analyzing public sentiment and pandemic trends. The proposed method is able to retrieve Twitter data related to COVID-19 and conduct spatiotemporal analysis. Furthermore, a deep learning component of the system provides monitoring and modeling capabilities for topics based on advanced natural language processing models. A variety of visualization methods are applied to the project to show the distribution of each topic. Our proposed system accurately reflects how public reactions change over time ...
The prompt estimation of traffic incident impacts can guide commuters in their trip planning and ... more The prompt estimation of traffic incident impacts can guide commuters in their trip planning and improve the resilience of transportation agencies' decision-making on resilience. However, it is more challenging than node-level and graphlevel forecasting tasks, as it requires extracting the anomaly subgraph or sub-time-series from dynamic graphs. In this paper, we propose DG-Trans, a novel traffic incident impact prediction framework, to foresee the impact of traffic incidents through dynamic graph learning. The proposed framework contains a dual-level spatial transformer and an importance-score-based temporal transformer, and the performance of this framework is justified by two newly constructed benchmark datasets. The duallevel spatial transformer removes unnecessary edges between nodes to isolate the affected subgraph from the other nodes. Meanwhile, the importance-score-based temporal transformer identifies abnormal changes in node features, causing the predictions to rely more on measurement changes after the incident occurs. Therefore, DG-Trans is equipped with dual abilities that extract spatiotemporal dependency and identify anomaly nodes affected by incidents while removing noise introduced by benign nodes. Extensive experiments on real-world datasets verify that DG-Trans outperforms the existing state-of-the-art methods, especially in extracting spatiotemporal dependency patterns and predicting traffic accident impacts. It offers promising potential for traffic incident management systems. Index Terms-spatiotemporal data mining, intelligent transportation systems, incident impact forecasting, transformer, anomaly detection.
Proceedings of the AAAI Conference on Artificial Intelligence
Forecasting citations of scientific patents and publications is a crucial task for understanding ... more Forecasting citations of scientific patents and publications is a crucial task for understanding the evolution and development of technological domains and for foresight into emerging technologies. By construing citations as a time series, the task can be cast into the domain of temporal point processes. Most existing work on forecasting with temporal point processes, both conventional and neural network-based, only performs single-step forecasting. In citation forecasting, however, the more salient goal is n-step forecasting: predicting the arrival time and the technology class of the next n citations. In this paper, we propose Dynamic Multi-Context Attention Networks (DMA-Nets), a novel deep learning sequence-to-sequence (Seq2Seq) model with a novel hierarchical dynamic attention mechanism for long-term citation forecasting. Extensive experiments on two real-world datasets demonstrate that the proposed model learns better representations of conditional dependencies over historical...
Proceedings of the AAAI Conference on Artificial Intelligence
Predicting and quantifying the impact of traffic accidents is necessary and critical to Intellige... more Predicting and quantifying the impact of traffic accidents is necessary and critical to Intelligent Transport Systems (ITS). As a state-of-the-art technique in graph learning, current graph neural networks heavily rely on graph Fourier transform, assuming homophily among the neighborhood. However, the homophily assumption makes it challenging to characterize abrupt signals such as traffic accidents. Our paper proposes an abrupt graph wavelet network (AGWN) to model traffic accidents and predict their time durations using only one single snapshot.
Proceedings of the AAAI Conference on Artificial Intelligence
Chinese characters have semantic-rich compositional information in radical form. While almost all... more Chinese characters have semantic-rich compositional information in radical form. While almost all previous research has applied CNNs to extract this compositional information, our work utilizes deep graph learning on a compact, graph-based representation of Chinese characters. This allows us to exploit temporal information within the strict stroke order used in writing characters. Our results show that our stroke-based model has potential for helping large-scale language models on some Chinese natural language understanding tasks. In particular, we demonstrate that our graph model produces more interpretable embeddings shown through word subtraction analogies and character embedding visualizations.
Proceedings of the AAAI Conference on Artificial Intelligence
Influence blocking maximization (IBM) is crucial in many critical real-world problems such as rum... more Influence blocking maximization (IBM) is crucial in many critical real-world problems such as rumors prevention and epidemic containment. The existing work suffers from: (1) concentrating on uniform costs at the individual level, (2) mostly utilizing greedy approaches to approximate optimization, (3) lacking a proper graph representation for influence estimates. To address these issues, this research introduces a neural network model dubbed Neural Influence Blocking (\algo) for improved approximation and enhanced influence blocking effectiveness. The code is available at https://github.com/oates9895/NIB.
2020 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM)
Open and accessible public utilities such as mass public transit systems are some of the vexing v... more Open and accessible public utilities such as mass public transit systems are some of the vexing venues that are vulnerable to several criminal acts due to the large volumes of commuters. Existing forms of threat or event detection for the rail-based transit systems are either not working in real-time or do not provide complete coverage. In this paper, we present RISECURE 1 , an open-source system, that uses real-time social media mining to aid in the early detection of such possible events within a rail-based/metro system. The system leverages dynamic query expansion to keep track of any new emerging information about any particular incident. The Real Time Incident panel of the proposed system provides a comprehensible representation of the evolution of threatening transit events, which are further shown in the storyline modal for each respective station. The alert notification module of the system is capable of monitoring threats to the rail-based/metro systems in real-time. We demonstrate the system by including case studies involving incidents occurring within the Washington DC Metropolitan Area Transit Authority (WMATA) metro system to justify the effectiveness of our approach.
Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence
Modeling and forecasting forward citations to a patent is a central task for the discovery of eme... more Modeling and forecasting forward citations to a patent is a central task for the discovery of emerging technologies and for measuring the pulse of inventive progress. Conventional methods for forecasting these forward citations cast the problem as analysis of temporal point processes which rely on the conditional intensity of previously received citations. Recent approaches model the conditional intensity as a chain of recurrent neural networks to capture memory dependency in hopes of reducing the restrictions of the parametric form of the intensity function. For the problem of patent citations, we observe that forecasting a patent's chain of citations benefits from not only the patent's history itself but also from the historical citations of assignees and inventors associated with that patent. In this paper, we propose a sequence-to-sequence model which employs an attention-of-attention mechanism to capture the dependencies of these multiple time sequences. Furthermore, th...
is paper presents SAFEBIKE, a novel route recommendation system for bike-sharing service that uti... more is paper presents SAFEBIKE, a novel route recommendation system for bike-sharing service that utilizes station information to infer the number of available bikes in dock and recommend bike routes according to multiple factors such as distance and safety level. e system consists of a station level availability predictor that predicts bikes and docks amount at each station, and an efcient route recommendation service that considers safety and bike/dock availability factors. It targets users who are concerned about route safeness and station availability. We demonstrate the system by utilizing Citi Bike station availability and New York City crime data of Manha an to show the e ectiveness of our approach. Integrated with real-time station availability and historical crime data resources, our proposed system can e ectively recommend an optimal bike route and improve the travel experience of bike users.
2018 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM), 2018
With the rapid growth in urban transit networks in recent years, detecting service disruptions in... more With the rapid growth in urban transit networks in recent years, detecting service disruptions in a timely manner is a problem of increased interest to service providers. Transit agencies are seeking to move beyond traditional customer questionnaires and manual service inspections to leveraging open source indicators like social media for deteting emerging transit events. In this paper, we leverage Twitter data for early detection of metro service disruptions. Inspired by the multi-task learning framework, we propose the Metro Disruption Detection Model, which captures the semantic similarity between transit lines in Twitter space. We propose novel constraints on feature semantic similarity exploiting prior knowledge about the spatial connectivity and shared tracks of the metro network. An algorithm based on the alternating direction method of multipliers (ADMM) framework is developed to solve the proposed model. We run extensive experiments and comparisons to other models with real world Twitter data and transit disruption records from the Washington Metropolitan Area Transit Authority (WMATA) to justify the efficacy of our model.
A Hierarchical Attention Graph Convolutional Network for Traffic Incident Impact Forecasting
2021 IEEE International Conference on Big Data (Big Data), 2021
Predicting the impact of traffic i ncidents b ased on traffic s ensor d ata i s a n e ssential r ... more Predicting the impact of traffic i ncidents b ased on traffic s ensor d ata i s a n e ssential r esearch t opic i n t he fi eld of Intelligent Transportation Systems (ITS). Tackling the problem of estimating the durations of incidents from their early stages is a challenge due to the variable nature of such incidents and the complex structure of modern road networks. Existing studies on forecasting the incident duration from sensor data are mostly incapable of modeling 1) the spatiotemporal correlations of traffic s ensors a nd a rterial r oads a nd 2 ) t he hierarchical topology of the traffic sensor and road networks. In this paper, we propose the Hierarchical Attention-based Spatiotemporal Graph Convolutional Network model (HastGCN) to solve the incident duration forecasting problem by formulating the spatiotemporal correlation and traffic p atterns o n b oth t he s ensor l evel and the road level in their natural hierarchical manner. At the sensor level, we propose a spatiotemporal attention mechanism followed by graph convolutions to model the local correlations and patterns between traffic s ensors o n t he s ame a rterial road. At the road level, a connectivity-aware attention mechanism is designed to learn the global spatial relatedness between each arterial road. Traffic-condition a ware g raph c onvolutions are then applied to understand the target incident representation for the incident duration forecasting.
Proceedings of the 23rd SIGSPATIAL International Conference on Advances in Geographic Information Systems, 2015
Traditional statistical analysis on speed, volume, and occupancy has dominated the field of Arter... more Traditional statistical analysis on speed, volume, and occupancy has dominated the field of Arterial Incident Management Study (AIMS). However, few previous works have focused on investigating into the causality of the incidents. In this paper, we present ButterFly, a social media based arterial incident detection and analysis system. The proposed system is dedicated to identify the traffic incident from a novel perspective and discover causalities between traffic incidents. The main functionalities of the proposed system include: 1) Traffic incident detection based on userinput social media contents, 2) Transportation incidents storyline generation, and 3) Traffic incidents causalities analysis and visualization. We demonstrate the system by considering the Washington DC area as our experimental environment. ButterFly is targeted to provide effective and convenient real-time and historical traffic incidents analysis interfaces for transportation management agencies and academies. Our proposed system, integrated with multiple social media resources, can greatly broaden the visions for traffic incidents analysis.
Uploads
Papers by Kaiqun Fu