In the evolving landscape of online information, disinformation is a growing concern. A concept central to this challenge is the "data void", a situation where there is a lack of relevant information online regarding certain search terms. This creates an opportunity for misleading or false narratives to fill the gap, often influencing public perception. In this work, we present methods to track and mitigate data voids in Web search settings.

A data void is a gap in online information, providing an opportunity for the spread of disinformation or a data void exploit. We introduce lightweight measures to track the progress of data void exploits and mitigation efforts in two contexts: Web search and Knowledge Graph (KG) querying. We use case studies to demonstrate the viability of these measures as data void trackers in the Web search context. To tackle data voids, we introduce an adversarial game model involving two agents: a disinformer and a mitigator. Both agents insert content into the information ecosystem to have their narrative rank higher than their counterpart in search results. At every turn, each agent chooses which content to deploy within their resource constraints, mimicking real-world situations where different entities have varying levels of influence and access to resources. Using simulations of this game, we compare and evaluate different mitigation strategies to recommend ones that maximize mitigation impact while minimizing costs.

Model-driven policymaking for epidemic control is a challenging collaborative process. It begins when a team of public-health officials, epidemiologists, and economists construct a reasonably predictive disease model representative of the team’s region of interest as a function of its unique socio-economic and demographic characteristics. As the team considers possible interventions such as school closures, social distancing, vaccination drives, etc., they need to simultaneously model each intervention’s effect on disease spread and economic cost. The team then engages in an extensive what-if analysis process to determine a cost-effective policy: a schedule of when, where and how extensively each intervention should be applied. This policymaking process is often an iterative and laborious programming-intensive effort where parameters are introduced and refined, model and intervention behaviors are modified, and schedules changed. We have designed and developed EpiPolicy to support this effort.

Synner allows users to generate realistic-looking data. With Synner users can visually and declaratively specify properties of the dataset they wish to generate. Such properties include the domain, and statistical distribution of each field, and relationships between fields. User can also sketch custom distributions and relationships. Synner provides instant feedback on every user interaction by visualizing a preview of the generated data. It also suggests generation specifications from a few user-provided examples of data to generate, column labels and other user interactions. In this demonstration, we showcase Synner and summarize results from our evaluation of Synner's effectiveness at generating realistic-looking data.

Everyone needs to make decisions under uncertainty and with limited resources, e.g., an investor who is building a stock portfolio subject to an investment budget and a bounded risk tolerance. Doing this with current technology is hard. There is a disconnect between software tools for data management, stochastic predictive modeling (e.g., simulation of future stock prices), and optimization; this leads to cumbersome analytical workflows. Moreover, current methods do not scale. To handle a broad class of uncertainty models, analysts approximate the original stochastic optimization problem by a large deterministic optimization problem that incorporates many "scenarios", i.e., sample realizations of the uncertain data values. For large problems, a huge number of scenarios is required, often causing the solver to fail. We demonstrate sPaQLTooLs, a system for in-database specification and scalable solution of constrained optimization problems. The key ingredients are (i) a database-oriented specification of constrained stochastic optimization problems as "stochastic package queries" (SPQs), (ii) use of a Monte Carlo database to incorporate stochastic predictive models, and (iii) a new SummarySearch algorithm for scalably solving SPQs with approximation guarantees. In this demonstration, the attendees will experience first-hand the difficulty of manually constructing feasible and high-quality portfolios, using real-world stock market data. We will then demonstrate how SummarySearch can easily and efficiently help them find very good portfolios, while being orders of magnitude faster than prior methods.

Synner is a tool that helps users generate real-looking synthetic data by visually and declaratively specifying the properties of the dataset such as each field's statistical distribution, its domain, and its relationship to other fields. It provides instant feedback on every user interaction by updating multiple visualizations of the generated dataset and even suggests data generation specifications from a few user examples and interactions. Synner visually communicates the inherent randomness of statistical data generation. Our evaluation of Synner demonstrates its effectiveness at generating realistic data when compared with Mockaroo, a popular data generation tool, and with hired developers who coded data generation scripts for a fee.

We present Qetch, a tool where users freely sketch patterns on a scale-less canvas to query time series data without specifying query length or amplitude. We study how humans sketch time series patterns --- humans preserve visually salient perceptual features but often non-uniformly scale and locally distort a pattern --- and we develop a novel matching algorithm that accounts for human sketching errors. Qetch enables the easy construction of complex and expressive queries with two key features: regular expressions over sketches and relative positioning of sketches to query multiple time-aligned series. Through user studies, we demonstrate the effectiveness of Qetch's different interaction features. We also demonstrate the effectiveness of Qetch's matching algorithm compared to popular algorithms on targeted, and exploratory query-by-sketch search tasks on a variety of data sets.

Query-by-sketch tools allow users to sketch a pattern to search a time series database for matches. Prior work adopts a bottom-up design approach: the sketching interface is built to reflect the inner workings of popular matching algorithms like Dynamic time warping (DTW) or Euclidean distance (ED). We design Qetch, a query-by-sketch tool for time series data, top-down. Users freely sketch patterns on a scale-less canvas. By studying how humans sketch time series patterns we develop a matching algorithm that accounts for human sketching errors. Qetch's top-down design and novel matching algorithm enable the easy construction of expressive queries that include regular expressions over sketches and queries over multiple time series. Our demonstration showcases Qetch and summarizes results from our evaluation of Qetch's effectiveness.