Embedding human computation in systems for data analysis improves the quality of the analysis, but can significantly impact the end-to-end cost and performance of the system. Recent work in crowdsourcing systems attempts to optimize for performance, but focuses on single applications running homogeneous tasks. In this work, we introduce Cioppino, a system that accounts for human factors that can affect performance when running multiple applications in parallel. Cioppino uses a queueing model to represent the pool of workers, and leverages techniques for autoscaling used in cloud computing to adaptively adjust the pool size. Its model also accounts for worker abandonment, and automatically shifts workers between applications to improve performance and match workers with tasks they enjoy most. Our evaluation of Cioppino in simulation on traces extracted from a realtime crowd system running on Amazon’s Mechanical Turk demonstrates a 19X reduction in cost, a 20% increase in throughput, and a 2X increase in worker preference for assigned tasks as compared to state-of-the-art crowd management strategies.

This paper studies the trade-off between two different kinds of pure exploration: breadth versus depth. We focus on the most biased coin problem, asking how many total coin flips are required to identify a ``heavy'' coin from an infinite bag containing both ``heavy'' coins with mean $\theta_1 \in (0,1)$, and ``light" coins with mean $\theta_0 \in (0,\theta_1)$, where heavy coins are drawn from the bag with proportion $\alpha \in (0,1/2)$. When $\alpha,\theta_0,\theta_1$ are unknown, the key difficulty of this problem lies in distinguishing whether the two kinds of coins have very similar means, or whether heavy coins are just extremely rare. While existing solutions to this problem require some prior knowledge of the parameters $\theta_0,\theta_1,\alpha$, we propose an adaptive algorithm that requires no such knowledge yet still obtains near-optimal sample complexity guarantees. In contrast, we provide a lower bound showing that non-adaptive strategies require at least quadratically more samples. In characterizing this gap between adaptive and nonadaptive strategies, we make connections to anomaly detection and prove lower bounds on the sample complexity of differentiating between a single parametric distribution and a mixture of two such distributions.

The proliferation of massive datasets combined with the development of sophisticated analytical techniques has enabled a wide variety of novel applications such as improved product recommendations, automatic image tagging, and improved speech-driven interfaces. A major obstacle to supporting these predictive applications is the challenging and expensive process of identifying and training an appropriate predictive model. Recent efforts aiming to automate this process have focused on single node implementations and have assumed that model training itself is a black box, limiting their usefulness for applications driven by large-scale datasets. In this work, we build upon these recent efforts and propose an architecture for automatic machine learning at scale comprised of a cost-based cluster resource allocation estimator, advanced hyper-parameter tuning techniques, bandit resource allocation via runtime algorithm introspection, and physical optimization via batching and optimal resource allocation. The result is TuPAQ, a component of the MLbase system that automatically finds and trains models for a user's predictive application with comparable quality to those found using exhaustive strategies, but an order of magnitude more efficiently than the standard baseline approach. TuPAQ scales to models trained on Terabytes of data across hundreds of machines.

Crowdsourced workflows are used in research and industry to solve a variety of tasks. The databases community has used crowd workers in query operators/optimization and for tasks such as entity resolution. Such research utilizes microtasks where crowd workers are asked to answer simple yes/no or multiple choice questions with little training. Typically, microtasks are used with voting algorithms to combine redundant responses from multiple crowd workers to achieve result quality. Microtasks are powerful, but fail in cases where larger context (e.g., domain knowledge) or significant time investment is needed to solve a problem, for example in large-document structured data extraction. In this paper, we consider context-heavy data processing tasks that may require many hours of work, and refer to such tasks as macrotasks. Leveraging the infrastructure and worker pools of existing crowdsourcing platforms, we automate macrotask scheduling, evaluation, and pay scales. A key challenge in macrotask-powered work, however, is evaluating the quality of a worker's output, since ground truth is seldom available and redundancy-based quality control schemes are impractical. We present Argonaut, a framework that improves macrotask powered work quality using a hierarchical review. Argonaut uses a predictive model of worker quality to select trusted workers to perform review, and a separate predictive model of task quality to decide which tasks to review. Finally, Argonaut can identify the ideal trade-off between a single phase of review and multiple phases of review given a constrained review budget in order to maximize overall output quality. We evaluate an industrial use of Argonaut to power a structured data extraction pipeline that has utilized over half a million hours of crowd worker input to complete millions of macrotasks. We show that Argonaut can capture up to 118% more errors than random spot-check reviews in review budget-constrained environments with up to two review layers.

Analysts report spending upwards of 80% of their time on problems in data cleaning. The data cleaning process is inherently iterative, with evolving cleaning workflows that start with basic exploratory data analysis on small samples of dirty data, then refine analysis with more sophisticated/expensive cleaning operators (e.g., crowdsourcing), and finally apply the insights to a full dataset. While an analyst often knows at a logical level what operations need to be done, they often have to manage a large search space of physical operators and parameters. We present Wisteria, a system designed to support the iterative development and optimization of data cleaning workflows, especially ones that utilize the crowd. Wisteria separates logical operations from physical implementations, and driven by analyst feedback, suggests optimizations and/or replacements to the analyst's choice of physical implementation. We highlight research challenges in sampling, in-flight operator replacement, and crowdsourcing. We overview the system architecture and these techniques, then provide a demonstration designed to showcase how Wisteria can improve iterative data analysis and cleaning. The code is available at: http://www.sampleclean.org.

Data labeling is a necessary but often slow process that impedes the development of interactive systems for modern data analysis. Despite rising demand for manual data labeling, there is a surprising lack of work addressing its high and unpredictable latency. In this paper, we introduce CLAMShell, a system that speeds up crowds in order to achieve consistently low-latency data labeling. We offer a taxonomy of the sources of labeling latency and study several large crowd-sourced labeling deployments to understand their empirical latency profiles. Driven by these insights, we comprehensively tackle each source of latency, both by developing novel techniques such as straggler mitigation and pool maintenance and by optimizing existing methods such as crowd retainer pools and active learning. We evaluate CLAMShell in simulation and on live workers on Amazon's Mechanical Turk, demonstrating that our techniques can provide an order of magnitude speedup and variance reduction over existing crowdsourced labeling strategies.