The rapidly evolving landscape of products, surfaces, policies, and regulations poses significant challenges for deploying state-of-the-art recommendation models at industry scale, primarily due to data fragmentation across domains and escalating infrastructure costs that hinder sustained quality improvements.

Abstract Task‐incremental learning (Task‐IL) aims to enable an intelligent agent to continuously accumulate knowledge from new learning tasks without catastrophically forgetting what it has learned in the past. It has drawn increasing attention in recent years, with many algorithms being proposed to mitigate neural network forgetting. However, none of the existing strategies is able to completely eliminate the issues. Moreover, explaining and fully understanding what knowledge and how it is being forgotten during the incremental learning process still remains under‐explored. In this paper, we propose KnowledgeDrift, a visual analytics framework, to interpret the network forgetting with three objectives: (1) to identify when the network fails to memorize the past knowledge, (2) to visualize what information has been forgotten, and (3) to diagnose how knowledge attained in the new model interferes with the one learned in the past. Our analytical framework first identifies the occurrence of forgetting by tracking the task performance under the incremental learning process and then provides in‐depth inspections of drifted information via various levels of data granularity. KnowledgeDrift allows analysts and model developers to enhance their understanding of network forgetting and compare the performance of different incremental learning algorithms. Three case studies are conducted in the paper to further provide insights and guidance for users to effectively diagnose catastrophic forgetting over time.

Deep learning based latent representations have been widely used for numerous scientific visualization applications such as isosurface similarity analysis, volume rendering, flow field synthesis, and data reduction, just to name a few. However, existing latent representations are mostly generated from raw data in an unsupervised manner, which makes it difficult to incorporate domain interest to control the size of the latent representations and the quality of the reconstructed data. In this paper, we present a novel importance-driven latent representation to facilitate domain-interest-guided scientific data visualization and analysis. We utilize spatial importance maps to represent various scientific interests and take them as the input to a feature transformation network to guide latent generation. We further reduced the latent size by a lossless entropy encoding algorithm trained together with the autoencoder, improving the storage and memory efficiency. We qualitatively and quantitatively evaluate the effectiveness and efficiency of latent representations generated by our method with data from multiple scientific visualization applications.

We propose VDL-Surrogate, a view-dependent neural-network-latent-based surrogate model for parameter space exploration of ensemble simulations that allows high-resolution visualizations and user-specified visual mappings. Surrogate-enabled parameter space exploration allows domain scientists to preview simulation results without having to run a large number of computationally costly simulations. Limited by computational resources, however, existing surrogate models may not produce previews with sufficient resolution for visualization and analysis. To improve the efficient use of computational resources and support high-resolution exploration, we perform ray casting from different viewpoints to collect samples and produce compact latent representations. This latent encoding process reduces the cost of surrogate model training while maintaining the output quality. In the model training stage, we select viewpoints to cover the whole viewing sphere and train corresponding VDL-Surrogate models for the selected viewpoints. In the model inference stage, we predict the latent representations at previously selected viewpoints and decode the latent representations to data space. For any given viewpoint, we make interpolations over decoded data at selected viewpoints and generate visualizations with user-specified visual mappings. We show the effectiveness and efficiency of VDL-Surrogate in cosmological and ocean simulations with quantitative and qualitative evaluations. Source code is publicly available at https://github.com/trainsn/VDL-Surrogate.

We present a novel technique for hierarchical super resolution (SR) with neural networks (NNs), which upscales volumetric data represented with an octree data structure to a high-resolution uniform gridwith minimal seam artifacts on octree node boundaries. Our method uses existing state-of-the-art SR models and adds flexibility to upscale input data with varying levels of detail across the domain, instead of only uniform grid data that are supported in previous approaches.The key is to use a hierarchy of SR NNs, each trained to perform 2× SR between two levels of detail, with a hierarchical SR algorithm that minimizes seam artifacts by starting from the coarsest level of detail and working up.We show that our hierarchical approach outperforms baseline interpolation and hierarchical upscaling methods, and demonstrate the usefulness of our proposed approach across three use cases including data reduction using hierarchical downsampling+SR instead of uniform downsampling+SR, computation savings for hierarchical finite-time Lyapunov exponent field calculation, and super-resolving low-resolution simulation results for a high-resolution approximation visualization.

We explore an online reinforcement learning (RL) paradigm to dynamically optimize parallel particle tracing performance in distributed-memory systems. Our method combines three novel components: (1) a work donation algorithm, (2) a high-order workload estimation model, and (3) a communication cost model. First, we design an RL-based work donation algorithm. Our algorithm monitors workloads of processes and creates RL agents to donate data blocks and particles from high-workload processes to low-workload processes to minimize program execution time. The agents learn the donation strategy on the fly based on reward and cost functions designed to consider processes' workload changes and data transfer costs of donation actions. Second, we propose a workload estimation model, helping RL agents estimate the workload distribution of processes in future computations. Third, we design a communication cost model that considers both block and particle data exchange costs, helping RL agents make effective decisions with minimized communication costs. We demonstrate that our algorithm adapts to different flow behaviors in large-scale fluid dynamics, ocean, and weather simulation data. Our algorithm improves parallel particle tracing performance in terms of parallel efficiency, load balance, and costs of I/O and communication for evaluations with up to 16,384 processors.

We propose GNN-Surrogate, a graph neural network-based surrogate model to explore the parameter space of ocean climate simulations. Parameter space exploration is important for domain scientists to understand the influence of input parameters (e.g., wind stress) on the simulation output (e.g., temperature). The exploration requires scientists to exhaust the complicated parameter space by running a batch of computationally expensive simulations. Our approach improves the efficiency of parameter space exploration with a surrogate model that predicts the simulation outputs accurately and efficiently. Specifically, GNN-Surrogate predicts the output field with given simulation parameters so scientists can explore the simulation parameter space with visualizations from user-specified visual mappings. Moreover, our graph-based techniques are designed for unstructured meshes, making the exploration of simulation outputs on irregular grids efficient. For efficient training, we generate hierarchical graphs and use adaptive resolutions. We give quantitative and qualitative evaluations on the MPAS-Ocean simulation to demonstrate the effectiveness and efficiency of GNN-Surrogate. Source code is publicly available at https://github.com/trainsn/GNN-Surrogate.

Viscous and gravitational flow instabilities cause a displacement front to break up into finger-like fluids. The detection and evolutionary analysis of these fingering instabilities are critical in multiple scientific disciplines such as fluid mechanics and hydrogeology. However, previous detection methods of the viscous and gravitational fingers are based on density thresholding, which provides limited geometric information of the fingers. The geometric structures of fingers and their evolution are important yet little studied in the literature. In this article, we explore the geometric detection and evolution of the fingers in detail to elucidate the dynamics of the instability. We propose a ridge voxel detection method to guide the extraction of finger cores from three-dimensional (3D) scalar fields. After skeletonizing finger cores into skeletons, we design a spanning tree based approach to capture how fingers branch spatially from the finger skeletons. Finally, we devise a novel geometric-glyph augmented tracking graph to study how the fingers and their branches grow, merge, and split over time. Feedback from earth scientists demonstrates the usefulness of our approach to performing spatio-temporal geometric analyses of fingers.

With the high growth rate of text data, extracting meaningful information from a large corpus becomes increasingly difficult. Keyword extraction and analysis is a common approach to tackle the problem, but it is non-trivial to identify important words in the text and represent the multifaceted properties of those words effectively. Traditional topic modeling based keyword analysis algorithms require hyper-parameters which are often difficult to tune without enough prior knowledge. In addition, the relationships among the keywords are often difficult to obtain. In this paper, we utilize the attention scores extracted from Transformer-based language models to capture word relationships. We propose a domain-driven attention tuning method, guiding the attention to learn domain-specific word relationships. From the attention, we build a keyword network and propose a novel algorithm, Attention-based Word Influence (AWI), to compute how influential each word is in the network. An interactive visual analytics system, KeywordMap, is developed to support multi-level analysis of keywords and keyword relationships through coordinated views. We measure the quality of keywords captured by our AWI algorithm quantitatively. We also evaluate the usefulness and effectiveness of KeywordMap through case studies.

We present the Feature Tracking Kit (FTK), a framework that simplifies, scales, and delivers various feature-tracking algorithms for scientific data. The key of FTK is our simplicial spacetime meshing scheme that generalizes both regular and unstructured spatial meshes to spacetime while tessellating spacetime mesh elements into simplices. The benefits of using simplicial spacetime meshes include (1) reducing ambiguity cases for feature extraction and tracking, (2) simplifying the handling of degeneracies using symbolic perturbations, and (3) enabling scalable and parallel processing. The use of simplicial spacetime meshing simplifies and improves the implementation of several feature-tracking algorithms for critical points, quantum vortices, and isosurfaces. As a software framework, FTK provides end users with VTK/ParaView filters, Python bindings, a command line interface, and programming interfaces for feature-tracking applications. We demonstrate use cases as well as scalability studies through both synthetic data and scientific applications including tokamak, fluid dynamics, and superconductivity simulations. We also conduct end-to-end performance studies on the Summit supercomputer. FTK is open sourced under the MIT license: https://github.com/hguo/ftk.

We present a novel distributed union-find algorithm that features asynchronous parallelism and k-d tree based load balancing for scalable visualization and analysis of scientific data. Applications of union-find include level set extraction and critical point tracking, but distributed union-find can suffer from high synchronization costs and imbalanced workloads across parallel processes. In this study, we prove that global synchronizations in existing distributed union-find can be eliminated without changing final results, allowing overlapped communications and computations for scalable processing. We also use a k-d tree decomposition to redistribute inputs, in order to improve workload balancing. We benchmark the scalability of our algorithm with up to 1,024 processes using both synthetic and application data. We demonstrate the use of our algorithm in critical point tracking and super-level set extraction with high-speed imaging experiments and fusion plasma simulations, respectively.

Cosmological N-body simulation is an important tool for scientists to study the evolution of the universe. With the increase of computing power, billions of particles of high space-time fidelity can be simulated by supercomputers. However, limited computer storage can only hold a small subset of the simulation output for analysis, which makes the understanding of the underlying cosmological phenomena difficult. To alleviate the problem, we design an in-situ data reduction method for large-scale unstructured particle data. During the data generation phase, we use a combined k-dimensional partitioning and Gaussian mixture model approach to reduce the data by utilizing probability distributions. We offer a model evaluation criterion to examine the quality of the probabilistic distribution models, which allows us to identify and improve low-quality models. After the in-situ processing, the particle data size is greatly reduced, which satisfies the requirements from the domain experts. By comparing the astronomical attributes and visualizations of the reconstructed data with the raw data, we demonstrate the effectiveness of our in-situ particle data reduction technique.

Cosmologists build simulations for the evolution of the universe using different initial parameters. By exploring the datasets from different simulation runs, cosmologists can understand the evolution of our universe and approach its initial conditions. A cosmological simulation nowadays can generate datasets on the order of petabytes. Moving datasets from the supercomputers to post data analysis machines is infeasible. We propose a novel approach called statistical super-resolution to tackle the big data problem for cosmological data analysis and visualization. It uses datasets from a few simulation runs to create a prior knowledge, which captures the relation between low-and high-resolution data. We apply in situ statistical down-sampling to datasets generated from simulation runs to minimize the requirements of I/O bandwidth and storage. High-resolution datasets are reconstructed from the statistical down-sampled data by using the prior knowledge for scientists to perform advanced data analysis and render high-quality visualizations.

Support vector machines (SVMs) are supervised learning models traditionally employed for classification and regression analysis. In classification analysis, a set of training data is chosen, and each instance in the training data is assigned a categorical class. An SVM then constructs a model based on a separating plane that maximizes the margin between different classes. Despite being one of the most popular classification models because of its strong performance empirically, understanding the knowledge captured in an SVM remains difficult. SVMs are typically applied in a black-box manner where the details of parameter tuning, training, and even the final constructed model are hidden from the users. This is natural since these details are often complex and difficult to understand without proper visualization tools. However, such an approach often brings about various problems including trial-and-error tuning and suspicious users who are forced to trust these models blindly. The contribution of this paper is a visual analysis approach for building SVMs in an open-box manner. Our goal is to improve an analyst’s understanding of the SVM modeling process through a suite of visualization techniques that allow users to have full interactive visual control over the entire SVM training process. Our visual exploration tools have been developed to enable intuitive parameter tuning, training data manipulation, and rule extraction as part of the SVM training process. To demonstrate the efficacy of our approach, we conduct a case study using a real-world robot control dataset.

The advanced telecom technologies and massive volumes of intelligent mobile phone users have yielded a huge amount of real-time data of people’s all-in-one telecommunication records, which we call telco big data. With telco data and the domain knowledge of an urban city, we are now able to analyze the movement and contact patterns of humans in an unprecedented scale. Flow map is widely used to display the movements of humans from one single source to multiple destinations by representing locations as nodes and movements as edges. However, it fails the task of visualizing both movement and contact data. In addition, analysts often need to compare and examine the patterns side by side, and do various quantitative analysis. In this work, we propose a novel spatio-temporal flow map layout to visualize when and where people from different locations move into the same places and make contact. We also propose integrating the spatiotemporal flow maps into existing spatiotemporal visualization techniques to form a suite of techniques for visualizing the movement and contact patterns. We report a potential application the proposed techniques can be applied to. The results show that our design and techniques properly unveil hidden information, while analysis can be achieved efficiently.

Understanding co-occurrence in urban human mobility (i.e. people from two regions visit an urban place during the same time span) is of great value in a variety of applications, such as urban planning, business intelligence, social behavior analysis, as well as containing contagious diseases. In recent years, the widespread use of mobile phones brings an unprecedented opportunity to capture large-scale and fine-grained data to study co-occurrence in human mobility. However, due to the lack of systematic and efficient methods, it is challenging for analysts to carry out in-depth analyses and extract valuable information. In this paper, we present TelCoVis, an interactive visual analytics system, which helps analysts leverage their domain knowledge to gain insight into the co-occurrence in urban human mobility based on telco data. Our system integrates visualization techniques with new designs and combines them in a novel way to enhance analysts' perception for a comprehensive exploration. In addition, we propose to study the correlations in co-occurrence (i.e. people from multiple regions visit different places during the same time span) by means of biclustering techniques that allow analysts to better explore coordinated relationships among different regions and identify interesting patterns. The case studies based on a real-world dataset and interviews with domain experts have demonstrated the effectiveness of our system in gaining insights into co-occurrence and facilitating various analytical tasks.