CHI ’25, Yokohama, Japan

Domain-specific languages (DSLs) for visualization aim to facilitate visualization creation by providing abstractions that offload implementation and execution details from users to the system layer. Therefore, DSLs often execute user-defined specifications by transforming them into intermediate representations (IRs) in successive lowering operations.

The success of deep learning has sparked interest in improving relational table tasks, like data preparation and search, with table representation models trained on large table corpora. Existing table corpora primarily contain tables extracted from HTML pages, limiting the capability to represent offline database tables. To train and evaluate high-capacity models for applications beyond the Web, we need resources with tables that resemble relational database tables. Here we introduce GitTables, a corpus of 1M relational tables extracted from GitHub. Our continuing curation aims at growing the corpus to at least 10M tables. Analyses of GitTables show that its structure, content, and topical coverage differ significantly from existing table corpora. We annotate table columns in GitTables with semantic types, hierarchical relations and descriptions from Schema.org and DBpedia. The evaluation of our annotation pipeline on the T2Dv2 benchmark illustrates that our approach provides results on par with human annotations. We present three applications of GitTables, demonstrating its value for learned semantic type detection models, schema completion methods, and benchmarks for table-to-KG matching, data search, and preparation. We make the corpus and code available at https://gittables.github.io.

Many data extraction tasks of practical relevance require not only syntactic pattern matching but also semantic reasoning about the content of the underlying text. While regular expressions are very well suited for tasks that require only syntactic pattern matching, they fall short for data extraction tasks that involve both a syntactic and semantic component. To address this issue, we introduce semantic regexes, a generalization of regular expressions that facilitates combined syntactic and semantic reasoning about textual data. We also propose a novel learning algorithm that can synthesize semantic regexes from a small number of positive and negative examples. Our proposed learning algorithm uses a combination of neural sketch generation and compositional type-directed synthesis for fast and effective generalization from a small number of examples. We have implemented these ideas in a new tool called Smore and evaluated it on representative data extraction tasks involving several textual datasets. Our evaluation shows that semantic regexes can better support complex data extraction tasks than standard regular expressions and that our learning algorithm significantly outperforms existing tools, including state-of-the-art neural networks and program synthesis tools.

Developers would benefit greatly from time travel: being able to faithfully replay past executions and retroactively execute modified code on past events. Currently, replay and retroaction are impractical because they require expensively capturing fine-grained timing information to reproduce concurrent accesses to shared state. In this paper, we propose practical time travel for database-backed applications , an important class of distributed applications that access shared state through transactions. We present R 3 , a novel Record-Replay-Retroaction tool. R 3 implements a lightweight interceptor to record concurrency information for applications at transaction-level granularity, enabling replay and retroaction with minimal overhead. We address key challenges in both replay and retroaction. First, we design a novel algorithm for faithfully reproducing application requests running with snapshot isolation, allowing R 3 to support most production DBMSs. Second, we develop a retroactive execution mechanism that provides high fidelity with the original trace while supporting nearly arbitrary code modifications. We demonstrate how R 3 simplifies debugging for real, hard-to-reproduce concurrency bugs from popular open-source web applications. We evaluate R 3 using TPC-C and microservice workloads and show that R 3 always-on recording has a small performance overhead (<25% for point queries but <0.1% for complex transactions like in TPC-C) during normal application execution and that R 3 can retroactively execute bugfixed code over recorded traces within 0.11--0.78× of the original execution time.

Icon arrays are graphical displays in which a subset of identical shapes are filled to convey probabilities. They are widely used for communicating probabilities to the general public. A primary design decision concerning icon arrays is how to fill and arrange these shapes. For example, a designer could fill the shapes from top to bottom or in a random fashion. We investigated the effect of different arrangements in icon arrays on probability perception. We showed participants icon arrays depicting probabilities between 0% and 100% in six different arrangements. Participants were more accurate in estimating probabilities when viewing the top, row, and diagonal arrangements, but they overestimated the proportions with the central arrangement and underestimated the proportions with the edge arrangement. They were biased to either overestimate or underestimate when viewing the random arrangement depending on the objective proportions, following a cyclical pattern consistent with existing findings in the psychophysics literature.

Inferring meta information about tables, such as column headers or relationships between columns, is an active research topic in data management as we find many tables are missing some of this information. In this paper, we study the problem of annotating table columns (i.e., predicting column types and the relationships between columns) using only information from the table itself. We develop a multi-task learning framework (called Doduo) based on pre-trained language models, which takes the entire table as input and predicts column types/relations using a single model. Experimental results show that Doduo establishes new state-of-the-art performance on two benchmarks for the column type prediction and column relation prediction tasks with up to 4.0% and 11.9% improvements, respectively. We report that Doduo can already outperform the previous state-of-the-art performance with a minimal number of tokens, only 8 tokens per column. We release a toolbox (https://github.com/megagonlabs/doduo) and confirm the effectiveness of Doduo on a real-world data science problem through a case study.

Cloud data warehouses (CDWs) bring large-scale data and compute power closer to users in enterprises. However, existing tools for analyzing data in CDWs are either limited in ad-hoc transformations or difficult to use for business users. Here we introduce Sigma Workbook, a new interactive system that enables business users to easily perform visual analysis of data in CDWs at scale. For this, Sigma Workbook provides an accessible spreadsheet-like interface for analysis through direct manipulation. Sigma Workbook dynamically constructs matching SQL queries from user interactions, building on the versatility and expressivity of SQL. Constructed queries are directly executed on CDWs, leveraging the superior characteristics of the new generation CDWs, including scalability. We demonstrate Sigma Workbook through 3 real-life use cases---cohort analysis, sessionization, and data augmentation---and underline Workbook's ease of use, scalability, and expressivity.

Reviews are integral to e-commerce services and products. They contain a wealth of information about the opinions and experiences of users, which can help better understand consumer decisions and improve user experience with products and services. Today, data scientists analyze reviews by developing rules and models to extract, aggregate, and understand information embedded in the review text. However, working with thousands of reviews, which are typically noisy incomplete text, can be daunting without proper tools. Here we first contribute results from an interview study that we conducted with fifteen data scientists who work with review text, providing insights into their practices and challenges. Results suggest data scientists need interactive systems for many review analysis tasks. Towards a solution, we then introduce Teddy, an interactive system that enables data scientists to quickly obtain insights from reviews and improve their extraction and modeling pipelines.

Researchers currently rely on ad hoc datasets to train automated visualization tools and evaluate the effectiveness of visualization designs. These exemplars often lack the characteristics of real-world datasets, and their one-off nature makes it difficult to compare different techniques. In this paper, we present VizNet: a large-scale corpus of over 31 million datasets compiled from open data repositories and online visualization galleries. On average, these datasets comprise 17 records over 3 dimensions and across the corpus, we find 51% of the dimensions record categorical data, 44% quantitative, and only 5% temporal. VizNet provides the necessary common baseline for comparing visualization design techniques, and developing benchmark models and algorithms for automating visual analysis. To demonstrate VizNet's utility as a platform for conducting online crowdsourced experiments at scale, we replicate a prior study assessing the influence of user task and data distribution on visual encoding effectiveness, and extend it by considering an additional task: outlier detection. To contend with running such studies at scale, we demonstrate how a metric of perceptual effectiveness can be learned from experimental results, and show its predictive power across test datasets.

Correctly detecting the semantic type of data columns is crucial for data science tasks such as automated data cleaning, schema matching, and data discovery. Existing data preparation and analysis systems rely on dictionary lookups and regular expression matching to detect semantic types. However, these matching-based approaches often are not robust to dirty data and only detect a limited number of types. We introduce Sherlock, a multi-input deep neural network for detecting semantic types. We train Sherlock on $686,765$ data columns retrieved from the VizNet corpus by matching $78$ semantic types from DBpedia to column headers. We characterize each matched column with $1,588$ features describing the statistical properties, character distributions, word embeddings, and paragraph vectors of column values. Sherlock achieves a support-weighted F$_1$ score of $0.89$, exceeding that of machine learning baselines, dictionary and regular expression benchmarks, and the consensus of crowdsourced annotations.

Abstract Pan and zoom are basic yet powerful interaction techniques for exploring large datasets. However, existing zoomable UI toolkits such as Pad++ and ZVTM do not provide the backend database support and data‐driven primitives that are necessary for creating large‐scale visualizations. This limitation in existing general‐purpose toolkits has led to many purpose‐built solutions (e.g. Google Maps and ForeCache) that address the issue of scalability but cannot be easily extended to support visualizations beyond their intended data types and usage scenarios. In this paper, we introduce Kyrix to ease the process of creating general and large‐scale web‐based pan/zoom visualizations. Kyrix is an integrated system that provides the developer with a concise and expressive declarative language along with a backend support for performance optimization of large‐scale data. To evaluate the scalability of Kyrix, we conducted a set of benchmarked experiments and show that Kyrix can support high interactivity (with an average latency of 100 ms or below) on pan/zoom visualizations of 100 million data points. We further demonstrate the accessibility of Kyrix through an observational study with 8 developers. Results indicate that developers can quickly learn Kyrix's underlying declarative model to create scalable pan/zoom visualizations. Finally, we provide a gallery of visualizations and show that Kyrix is expressive and flexible in that it can support the developer in creating a wide range of customized visualizations across different application domains and data types.

Visualizations of tabular data are widely used; understanding their effectiveness in different task and data contexts is fundamental to scaling their impact. However, little is known about how basic tabular data visualizations perform across varying data analysis tasks. In this paper, we report results from a crowdsourced experiment to evaluate the effectiveness of five small scale (5-34 data points) two-dimensional visualization types-Table, Line Chart, Bar Chart, Scatterplot, and Pie Chart-across ten common data analysis tasks using two datasets. We find the effectiveness of these visualization types significantly varies across task, suggesting that visualization design would benefit from considering context-dependent effectiveness. Based on our findings, we derive recommendations on which visualizations to choose based on different tasks. We finally train a decision tree on the data we collected to drive a recommender, showcasing how to effectively engineer experimental user data into practical visualization systems.

Dimensionality reduction is a common method for analyzing and visualizing high-dimensional data. However, reasoning dynamically about the results of a dimensionality reduction is difficult. Dimensionality-reduction algorithms use complex optimizations to reduce the number of dimensions of a dataset, but these new dimensions often lack a clear relation to the initial data dimensions, thus making them difficult to interpret. Here we propose a visual interaction framework to improve dimensionality-reduction based exploratory data analysis. We introduce two interaction techniques, forward projection and backward projection, for dynamically reasoning about dimensionally reduced data. We also contribute two visualization techniques, prolines and feasibility maps, to facilitate the effective use of the proposed interactions. We apply our framework to PCA and autoencoder-based dimensionality reductions. Through data-exploration examples, we demonstrate how our visual interactions can improve the use of dimensionality reduction in exploratory data analysis.

Abstract With the rapid commoditization of wearable sensors, detecting human movements from sensor datasets has become increasingly common over a wide range of applications. To detect activities, data scientists iteratively experiment with different classifiers before deciding which model to deploy. Effective reasoning about and comparison of alternative classifiers are crucial in successful model development. This is, however, inherently difficult in developing classifiers for sensor data, where the intricacy of long temporal sequences, high prediction frequency, and imprecise labeling make standard evaluation methods relatively ineffective and even misleading. We introduce Track Xplorer, an interactive visualization system to query, analyze, and compare the predictions of sensor‐data classifiers. Track Xplorer enables users to interactively explore and compare the results of different classifiers, and assess their accuracy with respect to the ground‐truth labels and video. Through integration with a version control system, Track Xplorer supports tracking of models and their parameters without additional workload on model developers. Track Xplorer also contributes an extensible algebra over track representations to filter, compose, and compare classification outputs, enabling users to reason effectively about classifier performance. We apply Track Xplorer in a collaborative project to develop classifiers to detect movements from multisensor data gathered from Parkinson's disease patients. We demonstrate how Track Xplorer helps identify early on possible systemic data errors, effectively track and compare the results of different classifiers, and reason about and pinpoint the causes of misclassifications.

Current tools for exploratory data analysis (EDA) require users to manually select data attributes, statistical computations and visual encodings. This can be daunting for large-scale, complex data. We introduce Foresight, a system that helps the user rapidly discover visual insights from large high-dimensional datasets. Formally, an "insight" is a strong manifestation of a statistical property of the data, e.g., high correlation between two attributes, high skewness or concentration about the mean of a single attribute, a strong clustering of values, and so on. For each insight type, Foresight initially presents visualizations of the top k instances in the data, based on an appropriate ranking metric. The user can then look at "nearby" insights by issuing "insight queries" containing constraints on insight strengths and data attributes. Thus the user can directly explore the space of insights, rather than the space of data dimensions and visual encodings as in other visual recommender systems. Foresight also provides "global" views of insight space to help orient the user and ensure a thorough exploration process. Furthermore, Foresight facilitates interactive exploration of large datasets through fast, approximate sketching.

In this paper we report the set-up and results of the Multimodal Brain Tumor Image Segmentation Benchmark (BRATS) organized in conjunction with the MICCAI 2012 and 2013 conferences. Twenty state-of-the-art tumor segmentation algorithms were applied to a set of 65 multi-contrast MR scans of low- and high-grade glioma patients-manually annotated by up to four raters-and to 65 comparable scans generated using tumor image simulation software. Quantitative evaluations revealed considerable disagreement between the human raters in segmenting various tumor sub-regions (Dice scores in the range 74%-85%), illustrating the difficulty of this task. We found that different algorithms worked best for different sub-regions (reaching performance comparable to human inter-rater variability), but that no single algorithm ranked in the top for all sub-regions simultaneously. Fusing several good algorithms using a hierarchical majority vote yielded segmentations that consistently ranked above all individual algorithms, indicating remaining opportunities for further methodological improvements. The BRATS image data and manual annotations continue to be publicly available through an online evaluation system as an ongoing benchmarking resource.

Visualization design can benefit from careful consideration of perception, as different assignments of visual encoding variables such as color, shape and size affect how viewers interpret data. In this work, we introduce perceptual kernels: distance matrices derived from aggregate perceptual judgments. Perceptual kernels represent perceptual differences between and within visual variables in a reusable form that is directly applicable to visualization evaluation and automated design. We report results from crowd-sourced experiments to estimate kernels for color, shape, size and combinations thereof. We analyze kernels estimated using five different judgment types--including Likert ratings among pairs, ordinal triplet comparisons, and manual spatial arrangement--and compare them to existing perceptual models. We derive recommendations for collecting perceptual similarities, and then demonstrate how the resulting kernels can be applied to automate visualization design decisions.

We introduce two-dimensional neural maps for exploring connectivity in the brain. For this, we create standard streamtube models from diffusion-weighted brain imaging data sets along with neural paths hierarchically projected into the plane. These planar neural maps combine desirable properties of low-dimensional representations, such as visual clarity and ease of tract-of-interest selection, with the anatomical familiarity of 3D brain models and planar sectional views. We distribute this type of visualization both in a traditional stand-alone interactive application and as a novel, lightweight web-accessible system. The web interface integrates precomputed neural-path representations into a geographical digital-maps framework with associated labels, metrics, statistics, and linkouts. Anecdotal and quantitative comparisons of the present method with a recently proposed 2D point representation suggest that our representation is more intuitive and easier to use and learn. Similarly, users are faster and more accurate in selecting bundles using the 2D path representation than the 2D point representation. Finally, expert feedback on the web interface suggests that it can be useful for collaboration as well as quick exploration of data.

We formulate the problem of surface deformations as the integration in the least square sense of a discrete vector-valued 1-forms obtained as the result of applying smooth stretching and rotation fields to the discrete differential of the 0-form defined by the vertex coordinates of a polygon mesh graph. Simple algorithms result from this formulation, which reduces to the solution of sparse linear systems. The method handles large angle rotations in one step and is invariant to rotations, translations, and scaling. We also introduce the integration of 1-forms along spanning trees as a heuristic to speed up the convergence of iterative solvers.

We introduce a new method for coloring 3D line fields and show results from its application in visualizing orientation in DTI brain data sets. The method uses Boy's surface, an immersion of RP2 in 3D. This coloring method is smooth and one-to-one except on a set of measure zero, the double curve of Boy's surface.

We present a visual exploration paradigm that facilitates navigation through complex fiber tracts by combining traditional 3D model viewing with lower dimensional representations. To this end, we create standard streamtube models along with two two-dimensional representations, an embedding in the plane and a hierarchical clustering tree, for a given set of fiber tracts. We then link these three representations using both interaction and color obtained by embedding fiber tracts into a perceptually uniform color space. We describe an anecdotal evaluation with neuroscientists to assess the usefulness of our method in exploring anatomical and functional structures in the brain. Expert feedback indicates that, while a standalone clinical use of the proposed method would require anatomical landmarks in the lower dimensional representations, the approach would be particularly useful in accelerating tract bundle selection. Results also suggest that combining traditional 3D model viewing with lower dimensional representations can ease navigation through the complex fiber tract models, improving exploration of the connectivity in the brain.

We present the results from a qualitative and quantitative user study comparing fishtank virtual-reality (VR) and CAVE displays. The results of the qualitative study show that users preferred the fishtank VR display to the CAVE system for our scientific visualization application because of perceived higher resolution, brightness and crispness of imagery, and comfort of use. The results of the quantitative study show that users performed an abstract visual search task significantly more quickly and more accurately on the fishtank VR display system than in the CAVE. The same study also showed that visual context had no significant effect on task performance for either of the platforms. We suggest that fishtank VR displays are more effective than CAVEs for applications in which the task occurs outside the user's reference frame, the user views and manipulates the virtual world from the outside in, and the size of the virtual object that the user interacts with is smaller than the user's body and fits into the fishtank VR display. The results of both studies support this proposition.

We present JointViewer, a software tool to aid orthopedics researchers in exploring complex, in-vivo joint kinematics. Given bone-geometry data and bone-motion information, JointViewer models and visualizes bone inter-spacing in the joint. Next, it proposes and displays plausible ligament paths which connect bones together. Both types of models are constructed through a distancefield approach. Users can maneuver the bones in a joint for better viewing, see motion relative to a specific bone, or remove bones from a joint. We demonstrate JointViewer’s effectiveness in three applications: examining normal human wrist kinematics, capturing the effect of injury on forearm kinematics, and exploring the kinematic constraints imposed by ligaments in a pigeon shoulder. In all applications, the system effectively highlights subtle yet important relationships among bones and soft-tissue that in previous standard joint visualizations had gone unnoticed.

We present a new method for visualizing 3D volumetric diffusion tensor MR images. We distinguish between linear anisotropy and planar anisotropy and represent values in the two regimes using streamtubes and streamsurfaces, respectively. Streamtubes represent structures with primarily linear diffusion, typically fiber tracts; streamtube direction correlates with tract orientation. The cross-sectional shape and color of each streamtube represent additional information from the diffusion tensor at each point. Streamsurfaces represent structures in which diffusion is primarily planar. Our algorithm chooses a very small representative subset of the streamtubes and streamsurfaces for display. We describe the set of metrics used for the culling process, which reduces visual clutter and improves interactivity. We also generate anatomical landmarks to identify the locations of such structures as the eyes, skull surface, and ventricles. The final models are complex surface geometries that can be imported into many interactive graphics software environments. We describe a virtual environment to interact with these models. Expert feedback from doctors studying changes in white-matter structures after gamma-knife capsulotomy and preoperative planning for brain tumor surgery shows that streamtubes correlate well with major neural structures, the 2D section and geometric landmarks are important in understanding the visualization, and the stereo and interactivity from the virtual environment aid in understanding the complex geometric models.

We describe a virtual reality environment for visualizing tensor-valued volumetric datasets acquired with diffusion tensor magnetic resonance imaging (DT-MRI). We have prototyped a virtual environment that displays geometric representations of the volumetric second-order diffusion tensor data and are developing interaction and visualization techniques for two application areas: studying changes in white-matter structures after gamma-knife capsulotomy and pre-operative planning for brain tumor surgery. Our feedback shows that compared to desktop displays, our system helps the user better interpret the large and complex geometric models, and facilitates communication among a group of users.