Mechanisms for generating differentially private synthetic data based on marginals and graphical models have been successful in a wide range of settings. However, one limitation of these methods is their inability to incorporate public data. Initializing a data generating model by pre-training on public data has shown to improve the quality of synthetic data, but this technique is not applicable when model structure is not determined a priori. We develop the mechanism JAM-PGM, which expands the adaptive measurements framework to jointly select between measuring public data and private data. This technique allows for public data to be included in a graphical-model-based mechanism. We show that JAM-PGM is able to outperform both publicly assisted and non publicly assisted synthetic data generation mechanisms even when the public data distribution is biased.

Differential privacy (DP) has the potential to enable privacy-preserving analysis on sensitive data, but requires analysts to judiciously spend a limited "privacy loss budget" ϵ across queries. Analysts conducting exploratory analyses do not, however, know all queries in advance and seldom have DP expertise. Thus, they are limited in their ability to specify ϵ allotments across queries prior to an analysis. To support analysts in spending ϵ efficiently, we propose a new interactive analysis paradigm, Measure-Observe-Remeasure, where analysts "measure" the database with a limited amount of ϵ, observe estimates and their errors, and remeasure with more ϵ as needed.We instantiate the paradigm in an interactive visualization interface which allows analysts to spend increasing amounts of ϵ under a total budget. To observe how analysts interact with the Measure-Observe-Remeasure paradigm via the interface, we conduct a user study that compares the utility of ϵ allocations and findings from sensitive data participants make to the allocations and findings expected of a rational agent who faces the same decision task. We find that participants are able to use the workflow relatively successfully, including using budget allocation strategies that maximize over half of the available utility stemming from ϵ allocation. Their loss in performance relative to a rational agent appears to be driven more by their inability to access information and report it than to allocate ϵ.

We propose AIM, a new algorithm for differentially private synthetic data generation. AIM is a workload-adaptive algorithm within the paradigm of algorithms that first selects a set of queries, then privately measures those queries, and finally generates synthetic data from the noisy measurements. It uses a set of innovative features to iteratively select the most useful measurements, reflecting both their relevance to the workload and their value in approximating the input data. We also provide analytic expressions to bound per-query error with high probability which can be used to construct confidence intervals and inform users about the accuracy of generated data. We show empirically that AIM consistently outperforms a wide variety of existing mechanisms across a variety of experimental settings.

Many differentially private algorithms for answering database queries involve astep that reconstructs a discrete data distribution from noisy measurements. Thisprovides consistent query answers and reduces error, but often requires space thatgrows exponentially with dimension. PRIVATE-PGM is a recent approach that usesgraphical models to represent the data distribution, with complexity proportional tothat of exact marginal inference in a graphical model with structure determined bythe co-occurrence of variables in the noisy measurements. PRIVATE-PGM is highlyscalable for sparse measurements, but may fail to run in high dimensions with densemeasurements. We overcome the main scalability limitation of PRIVATE-PGMthrough a principled approach that relaxes consistency constraints in the estimationobjective. Our new approach works with many existing private query answeringalgorithms and improves scalability or accuracy with no privacy cost.

Differential Privacy is an emerging privacy model with increasing popularity in many domains. It functions by adding carefully calibrated noise to data that blurs information about individuals while preserving overall statistics about the population. Theoretically, it is possible to produce robust privacy-preserving visualizations by plotting differentially private data. However, noise-induced data perturbations can alter visual patterns and impact the utility of a private visualization. We still know little about the challenges and opportunities for visual data exploration and analysis using private visualizations. As a first step towards filling this gap, we conducted a crowdsourced experiment, measuring participants' performance under three levels of privacy (high, low, non-private) for combinations of eight analysis tasks and four visualization types (bar chart, pie chart, line chart, scatter plot). Our findings show that for participants' accuracy for summary tasks (e.g., find clusters in data) was higher that value tasks (e.g., retrieve a certain value). We also found that under DP, pie chart and line chart offer similar or better accuracy than bar chart. In this work, we contribute the results of our empirical study, investigating the task-based effectiveness of basic private visualizations, a dichotomous model for defining and measuring user success in performing visual analysis tasks under DP, and a set of distribution metrics for tuning the injection to improve the utility of private visualizations.

The adoption of differential privacy is growing, but the complexity of designing private, efficient, and accurate algorithms is still high. We propose a novel programming framework and system, ϵ KTELO for implementing both existing and new privacy algorithms. For the task of answering linear counting queries, we show that nearly all existing algorithms can be composed from operators, each conforming to one of a small number of operator classes. While past programming frameworks have helped to ensure the privacy of programs, the novelty of our framework is its significant support for authoring accurate and efficient (as well as private) programs. After describing the design and architecture of the ϵ KTELO system, we show that ϵ KTELO is expressive, allows for safer implementations through code reuse, and allows both privacy novices and experts to easily design algorithms. We provide a number of novel implementation techniques to support the generality and scalability of ϵ KTELO operators. These include methods to automatically compute lossless reductions of the data representation, implicit matrices that avoid materialized state but still support computations, and iterative inference implementations that generalize techniques from the privacy literature. We demonstrate the utility of ϵ KTELO by designing several new state-of-the-art algorithms, most of which result from simple re-combinations of operators defined in the framework. We study the accuracy and scalability of ϵ KTELO plans in a thorough empirical evaluation.

Many privacy mechanisms reveal high-level information about a data distribution through noisy measurements. It is common to use this information to estimate the answers to new queries. In this work, we provide an approach to solve this estimation problem efficiently using graphical models, which is particularly effective when the distribution is high-dimensional but the measurements are over low-dimensional marginals. We show that our approach is far more efficient than existing estimation techniques from the privacy literature and that it can improve the accuracy and scalability of many state-of-the-art mechanisms.

Items from a database are often ranked based on a combination of criteria. The weight given to each criterion in the combination can greatly affect the ranking produced. Often, a user may have a general sense of the relative importance of the different criteria, but beyond this may have the flexibility, within limits, to choose combinations that weigh these criteria differently with an acceptable region. We demonstrate MithraRanking, a system that helps users choose criterion weights that lead to "better'' rankings in terms of having desirable properties while remaining within the acceptable region. The goodness properties we focus on are stability and fairness.

Across many application domains, trusted parties who collect sensitive information need mechanisms to safely disseminate data. A favored approach is to generate synthetic data : a dataset similar to the original, hopefully retaining its statistical features, but one that does not reveal the private information of contributors to the data. We present PSynDB, a web-based synthetic table generator that is built on recent privacy technologies [10,11,15]. PSynDB satisfies the formal guarantee of differential privacy and generates synthetic tables with high accuracy for tasks that the user specifies as important. PSynDB allows users to browse expected error rates before running the mechanism, a useful feature for making important policy decisions, such as setting the privacy loss budget. When the user has finished configuration, the tool outputs a data synthesis program that can be ported to a trusted environment. There it can be safely executed on the private data to produce the private synthetic dataset for broad dissemination.

Differential privacy is considered a de facto standard for private data analysis. However, the definition and much of the supporting literature applies to flat tables. While there exist variants of the definition and specialized algorithms for specific types of relational data (e.g. graphs), there isn't a general privacy definition for multi-relational schemas with constraints, and no system that permits accurate differentially private answering of SQL queries while imposing a fixed privacy budget across all queries posed by the analyst. This work presents PrivateSQL, a first-of-its-kind end-to-end differentially private relational database system. PrivateSQL allows an analyst to query data stored in a standard database management system using a rich class of SQL counting queries. PrivateSQL adopts a novel generalization of differential privacy to multi-relational data that takes into account constraints in the schema like foreign keys, and allows the data owner to flexibly specify entities in the schema that need privacy. PrivateSQL ensures a fixed privacy loss across all the queries posed by the analyst by answering queries on private synopses generated from several views over the base relation that are tuned to have low error on a representative query workload. We experimentally evaluate PrivateSQL on a real-world dataset and a workload of more than 3, 600 queries. We show that for 50% of the queries PrivateSQL offers at least 1, 000x better error rates than solutions adapted from prior work.

Emerging IoT technologies promise to bring revolutionary changes to many domains including health, transportation, and building management. However, continuous monitoring of individuals threatens privacy. The success of IoT thus depends on integrating privacy protections into IoT infrastructures. This demonstration adapts a recently-proposed system, PeGaSus, which releases streaming data under the formal guarantee of differential privacy, with a state-of-the-art IoT testbed (TIPPERS) located at UC Irvine. PeGaSus protects individuals' data by introducing distortion into the output stream. While PeGaSuS has been shown to offer lower numerical error compared to competing methods, assessing the usefulness of the output is application dependent.

Algorithmic decisions often result in scoring and ranking individuals to determine credit worthiness, qualifications for college admissions and employment, and compatibility as dating partners. While automatic and seemingly objective, ranking algorithms can discriminate against individuals and protected groups, and exhibit low diversity. Furthermore, ranked results are often unstable -- small changes in the input data or in the ranking methodology may lead to drastic changes in the output, making the result uninformative and easy to manipulate. Similar concerns apply in cases where items other than individuals are ranked, including colleges, academic departments, or products. Despite the ubiquity of rankers, there is, to the best of our knowledge, no technical work that focuses on making rankers transparent.

The adoption of differential privacy is growing but the complexity of designing private, efficient and accurate algorithms is still high. We propose a novel programming framework and system, Ektelo, for implementing both existing and new privacy algorithms. For the task of answering linear counting queries, we show that nearly all existing algorithms can be composed from operators, each conforming to one of a small number of operator classes. While past programming frameworks have helped to ensure the privacy of programs, the novelty of our framework is its significant support for authoring accurate and efficient (as well as private) programs. After describing the design and architecture of the Ektelo system, we show that Ektelo is expressive, that it allows for safer implementations through code reuse, and that it allows both privacy novices and experts to easily design algorithms. We demonstrate the use of Ektelo by designing several new state-of-the-art algorithms.

Decision making is challenging when there is more than one criterion to consider. In such cases, it is common to assign a goodness score to each item as a weighted sum of its attribute values and rank them accordingly. Clearly, the ranking obtained depends on the weights used for this summation. Ideally, one would want the ranked order not to change if the weights are changed slightly. We call this property stability of the ranking. A consumer of a ranked list may trust the ranking more if it has high stability. A producer of a ranked list prefers to choose weights that result in a stable ranking, both to earn the trust of potential consumers and because a stable ranking is intrinsically likely to be more meaningful. In this paper, we develop a framework that can be used to assess the stability of a provided ranking and to obtain a stable ranking within an "acceptable" range of weight values (called "the region of interest"). We address the case where the user cares about the rank order of the entire set of items, and also the case where the user cares only about the top- k items. Using a geometric interpretation, we propose algorithms that produce stable rankings. In addition to theoretical analyses, we conduct extensive experiments on real datasets that validate our proposal.

Differentially private algorithms for answering sets of predicate counting queries on a sensitive database have many applications. Organizations that collect individual-level data, such as statistical agencies and medical institutions, use them to safely release summary tabulations. However, existing techniques are accurate only on a narrow class of query workloads, or are extremely slow, especially when analyzing more than one or two dimensions of the data. In this work we propose HDMM, a new differentially private algorithm for answering a workload of predicate counting queries, that is especially effective for higher-dimensional datasets. HDMM represents query workloads using an implicit matrix representation and exploits this compact representation to efficiently search (a subset of) the space of differentially private algorithms for one that answers the input query workload with high accuracy. We empirically show that HDMM can efficiently answer queries with lower error than state-of-the-art techniques on a variety of low and high dimensional datasets.

Recently, there has begun a movement towards Fairness, Accountability, and Transparency (FAT) in algorithmic decision making, and in data science more broadly. The database community has not been significantly involved in this movement, despite "owning" the models, languages, and systems that produce the (potentially biased) input to the machine learning applications. What role should the database community play in this movement? Do the objectives of fairness, accountability and transparency give rise to core data management issues that can drive new research questions and new systems, or are these "soft topics" that are best left to be managed with policy? Will emphasis on these topics dilute our core competency in techniques and technologies for data, or can it reinforce our central role in technology stacks ranging from startups to the enterprise, and from local non-profits to the federal government? The goal of this panel is to debate these questions, and to whet the appetite of the data management community for research in this important emerging area.

Query result diversification is widely used in data exploration, Web search, and recommendation systems. The problem of returning diversified query results consists of finding a small subset of valid query answers that are representative and different from one another, usually quantified by a diversity score. Most existing techniques for query diversification first compute all valid query results and then find a diverse subset. These techniques are inefficient when the set of valid query results is large. Other work has proposed efficient solutions for restricted application settings, where results are shared across multiple queries. In this paper, our goal is to support result diversification for general range queries over a single relation. We propose the RC-Index, a novel index structure that achieves efficiency by reducing the number of items that must be retrieved by the database to form a diverse set of the desired size (about 1 second for a dataset of 1 million items). Further, we prove that an RC-Index offers strong approximation guarantees. To the best of our knowledge, this is the first index-based diversification method with a guaranteed approximation ratio for range queries.

Individuals are continually observed by an ever-increasing number of sensors that make up the Internet of Things. The resulting streams of data, which are analyzed in real time, can reveal sensitive personal information about individuals. Hence, there is an urgent need for stream processing solutions that can analyze these data in real time with provable guarantees of privacy and low error.

We investigate the problem of learning discrete graphical models in a differentially private way. Approaches to this problem range from privileged algorithms that conduct learning completely behind the privacy barrier to schemes that release private summary statistics paired with algorithms to learn parameters from those statistics. We show that the approach of releasing noisy sufficient statistics using the Laplace mechanism achieves a good trade-off between privacy, utility, and practicality. A naive learning algorithm that uses the noisy sufficient statistics “as is” outperforms general-purpose differentially private learning algorithms. However, it has three limitations: it ignores knowledge about the data generating process, rests on uncertain theoretical foundations, and exhibits certain pathologies. We develop a more principled approach that applies the formalism of collective graphical models to perform inference over the true sufficient statistics within an expectation-maximization framework. We show that this learns better models than competing approaches on both synthetic data and on real human mobility data used as a case study.

Given a collection of rankings of a set of items, rank aggregation seeks to compute a ranking that can serve as a single best representative of the collection. Rank aggregation is a well-studied problem and a number of effective algorithmic solutions have been proposed in the literature. However, when individuals are asked to contribute a ranking, they may be concerned that their personal preferences will be disclosed inappropriately to others. This acts as a disincentive to individuals to respond honestly in expressing their preferences and impedes data collection and data sharing. We address this problem by investigating rank aggregation under differential privacy, which requires that a released output (here, the aggregate ranking computed from individuals' rankings) remain almost the same if any one individual's ranking is removed from the input. We propose a number of differentially-private rank aggregation algorithms: two are inspired by non-private approximate rank aggregators from the existing literature; another uses a novel rejection sampling method to sample privately from a complex distribution. For all the methods we propose, we quantify, both theoretically and empirically, the “cost” of privacy in terms of the quality of the rank aggregation computed.

Differential privacy has emerged as the dominant privacy standard for data analysis. Its wide acceptance has led to significant development of algorithms that meet this rigorous standard. For some tasks, such as the task of answering low dimensional counting queries, dozens of algorithms have been proposed. However, no single algorithm has emerged as the dominant performer, and in fact, algorithm performance varies drastically across inputs. Thus, it's not clear how to select an algorithm for a particular task, and choosing the wrong algorithm might lead to significant degradation in terms of analysis accuracy. We believe that the difficulty of algorithm selection is one factor limiting the adoption of differential privacy in real systems. In this demonstration we present DIAS (Differentially-private Interactive Algorithm Selection), an educational privacy game. Users are asked to perform algorithm selection for a variety of inputs and compare the performance of their choices against that of Pythia, an automated algorithm selection framework. Our hope is that by the end of the game users will understand the importance of algorithm selection and most importantly will have a good grasp on how to use differentially private algorithms for their own applications.

Differential privacy has emerged as a preferred standard for ensuring privacy in analysis tasks on sensitive datasets. Recent algorithms have allowed for significantly lower error by adapting to properties of the input data. These so-called data-dependent algorithms have different error rates for different inputs. There is now a complex and growing landscape of algorithms without a clear winner that can offer low error over all datasets. As a result, the best possible error rates are not attainable in practice, because the data curator cannot know which algorithm to select prior to actually running the algorithm.

Issues of responsible data analysis and use are coming to the forefront of the discourse in data science research and practice, with most significant efforts to date on the part of the data mining, machine learning, and security and privacy communities. In these fields, the research has been focused on analyzing the fairness, accountability and transparency (FAT) properties of specific algorithms and their outputs. Although these issues are most apparent in the social sciences where fairness is interpreted in terms of the distribution of resources across protected groups, management of bias in source data affects a variety of fields. Consider climate change studies that require representative data from geographically diverse regions, or supply chain analyses that require data that represents the diversity of products and customers. Any domain that involves sparse or sampled data has exposure to potential bias.

Big data technology holds incredible promise of improving people's lives, accelerating scientific discovery and innovation , and bringing about positive societal change. Yet, if not used responsibly, this technology can propel economic inequality , destabilize global markets and affirm systemic bias. While the potential benefits of big data are well-accepted, the importance of using these techniques in a fair and transparent manner is rarely considered. The primary goal of this tutorial is to draw the attention of the data management community to the important emerging subject of responsible data management and analysis. We will offer our perspective on the issue, will give an overview of existing technical work, primarily from the data mining and algorithms communities, and will motivate future research directions.

Differential privacy has become the dominant standard in the research community for strong privacy protection. There has been a flood of research into query answering algorithms that meet this standard. Algorithms are becoming increasingly complex, and in particular, the performance of many emerging algorithms is data dependent, meaning the distribution of the noise added to query answers may change depending on the input data. Theoretical analysis typically only considers the worst case, making empirical study of average case performance increasingly important. In this paper we propose a set of evaluation principles which we argue are essential for sound evaluation. Based on these principles we propose DPBench, a novel evaluation framework for standardized evaluation of privacy algorithms. We then apply our benchmark to evaluate algorithms for answering 1- and 2-dimensional range queries. The result is a thorough empirical study of 15 published algorithms on a total of 27 datasets that offers new insights into algorithm behavior---in particular the influence of dataset scale and shape---and a more complete characterization of the state of the art. Our methodology is able to resolve inconsistencies in prior empirical studies and place algorithm performance in context through comparison to simple baselines. Finally, we pose open research questions which we hope will guide future algorithm design.

The emergence of differential privacy as a primary standard for privacy protection has led to the development, by the research community, of hundreds of algorithms for various data analysis tasks. Yet deployment of these techniques has been slowed by the complexity of algorithms and an incomplete understanding of the cost to accuracy implied by the adoption of differential privacy. In this demonstration we present DPComp, a publicly-accessible web-based system, designed to support a broad community of users, including data analysts, privacy researchers, and data owners. Users can use DPComp to assess the accuracy of state-of-the-art privacy algorithms and interactively explore algorithm output in order to understand, both quantitatively and qualitatively, the error introduced by the algorithms. In addition, users can contribute new algorithms and new (non-sensitive) datasets. DPComp automatically incorporates user contributions into an evolving benchmark based on a rigorous evaluation methodology articulated by Hay et al. (SIGMOD 2016).

The availability of public computing resources in the cloud has revolutionized data analysis, but requesting cloud resources often involves complex decisions for consumers. Estimating the completion time and cost of a computation and requesting the appropriate cloud resources are challenging tasks even for an expert user. We propose a new market-based framework for pricing computational tasks in the cloud. Our framework introduces an agent between consumers and cloud providers. The agent takes data and computational tasks from users, estimates time and cost for evaluating the tasks, and returns to consumers contracts that specify the price and completion time. Our framework can be applied directly to existing cloud markets without altering the way cloud providers offer and price services. In addition, it simplifies cloud use for consumers by allowing them to compare contracts, rather than choose resources directly. We present design, analytical, and algorithmic contributions focusing on pricing computation contracts, analyzing their properties, and optimizing them in complex workflows. We conduct an experimental evaluation of our market framework over a real-world cloud service and demonstrate empirically that our market ensures three key properties: (a) that consumers benefit from using the market due to competitiveness among agents, (b) that agents have an incentive to price contracts fairly, and (c) that inaccuracies in estimates do not pose a significant risk to agents' profits. Finally, we present a fine-grained pricing mechanism for complex workflows and show that it can increase agent profits by more than an order of magnitude in some cases.

The management of Web users' personal information is increasingly distributed across a broad array of applications and systems, including online social networks and cloud-based services. Users wish to share data using these systems, but avoiding the risks of unintended disclosures or unauthorized access by applications has become a major challenge.

Evaluating the performance of database systems is crucial when database vendors or researchers are developing new technologies. But such evaluation tasks rely heavily on actual data and query workloads that are often unavailable to researchers due to privacy restrictions. To overcome this barrier, we propose a framework for the release of a synthetic database which accurately models selected performance properties of the original database. We improve on prior work on synthetic database generation by providing a formal, rigorous guarantee of privacy. Accuracy is achieved by generating synthetic data using a carefully selected set of statistical properties of the original data which balance privacy loss with relevance to the given query workload. An important contribution of our framework is an extension of standard differential privacy to multiple tables.

The effective analysis of social networks and graph-structured data is often limited by the privacy concerns of individuals whose data make up these networks. Differential privacy offers individuals a rigorous and appealing guarantee of privacy. But while differentially private algorithms for computing basic graph properties have been proposed, most graph modeling tasks common in the data mining community cannot yet be carried out privately.

We describe a new algorithm for answering a given set of range queries under ε-differential privacy which often achieves substantially lower error than competing methods. Our algorithm satisfies differential privacy by adding noise that is adapted to the input data and to the given query set. We first privately learn a partitioning of the domain into buckets that suit the input data well. Then we privately estimate counts for each bucket, doing so in a manner well-suited for the given query set. Since the performance of the algorithm depends on the input database, we evaluate it on a wide range of real datasets, showing that we can achieve the benefits of data-dependence on both "easy" and "hard" databases.

Personal data has value to both its owner and to institutions who would like to analyze it. Privacy mechanisms protect the owner's data while releasing to analysts noisy versions of aggregate query results. But such strict protections of the individual's data have not yet found wide use in practice. Instead, Internet companies, for example, commonly provide free services in return for valuable sensitive information from users, which they exploit and sometimes sell to third parties. As awareness of the value of personal data increases, so has the drive to compensate the end-user for her private information. The idea of monetizing private data can improve over the narrower view of hiding private data, since it empowers individuals to control their data through financial means. In this article we propose a theoretical framework for assigning prices to noisy query answers as a function of their accuracy, and for dividing the price amongst data owners who deserve compensation for their loss of privacy. Our framework adopts and extends key principles from both differential privacy and query pricing in data markets. We identify essential properties of the pricing function and micropayments, and characterize valid solutions.

Personal data has value to both its owner and to institutions who would like to analyze it. Privacy mechanisms protect the owner's data while releasing to analysts noisy versions of aggregate query results. But such strict protections of individual's data have not yet found wide use in practice. Instead, Internet companies, for example, commonly provide free services in return for valuable sensitive information from users, which they exploit and sometimes sell to third parties.

A common goal of privacy research is to release synthetic data that satisfies a formal privacy guarantee and can be used by an analyst in place of the original data. To achieve reasonable accuracy, a synthetic data set must be tuned to support a specified set of queries accurately, sacrificing fidelity for other queries.

We present the WebdamLog system for managing distributed data on the Web in a peer-to-peer manner. We demonstrate the main features of the system through an application called Wepic for sharing pictures between attendees of the sigmod conference. Using Wepic, the attendees will be able to share, download, rate and annotate pictures in a highly decentralized manner. We show how WebdamLog handles heterogeneity of the devices and services used to share data in such a Web setting. We exhibit the simple rules that define the Wepic application and show how to easily modify the Wepic application.

Data privacy has been an important research topic in the security, theory and database communities in the last few decades. However, many existing studies have restrictive assumptions regarding the adversary's prior knowledge, meaning that they preserve individuals' privacy only when the adversary has rather limited background information about the sensitive data, or only uses certain kinds of attacks. Recently, differential privacy has emerged as a new paradigm for privacy protection with very conservative assumptions about the adversary's prior knowledge. Since its proposal, differential privacy had been gaining attention in many fields of computer science, and is considered among the most promising paradigms for privacy-preserving data publication and analysis. In this tutorial, we will motivate its introduction as a replacement for other paradigms, present the basics of the differential privacy model from a database perspective, describe the state of the art in differential privacy research, explain the limitations and shortcomings of differential privacy, and discuss open problems for future research.

We propose a novel mechanism for answering sets of counting queries under differential privacy. Given a workload of counting queries, the mechanism automatically selects a different set of "strategy" queries to answer privately, using those answers to derive answers to the workload. The main algorithm proposed in this paper approximates the optimal strategy for any workload of linear counting queries. With no cost to the privacy guarantee, the mechanism improves significantly on prior approaches and achieves near-optimal error for many workloads, when applied under (ε, δ)-differential privacy. The result is an adaptive mechanism which can help users achieve good utility without requiring that they reason carefully about the best formulation of their task.

The proliferation of information networks, as a means of sharing information, has raised privacy concerns for enterprises who manage such networks and for individual users that participate in such networks. For enterprises, the main challenge is to satisfy two competing goals: releasing network data for useful data analysis and also preserving the identities or sensitive relationships of the individuals participating in the network. Individual users, on the other hand, require personalized methods that increase their awareness of the visibility of their private information.

Differential privacy is a robust privacy standard that has been successfully applied to a range of data analysis tasks. But despite much recent work, optimal strategies for answering a collection of related queries are not known.

We show that it is possible to significantly improve the accuracy of a general class of histogram queries while satisfying differential privacy. Our approach carefully chooses a set of queries to evaluate, and then exploits consistency constraints that should hold over the noisy output. In a post-processing phase, we compute the consistent input most likely to have produced the noisy output. The final output is differentially-private and consistent, but in addition, it is often much more accurate. We show, both theoretically and experimentally, that these techniques can be used for estimating the degree sequence of a graph very precisely, and for computing a histogram that can support arbitrary range queries accurately.

Incorporating probabilities into the semantics of incomplete databases has posed many challenges, forcing systems to sacrifice modeling power, scalability, or treatment of relational algebra operators. We propose an alternative approach where the underlying relational database always represents a single world, and an external factor graph encodes a distribution over possible worlds; Markov chain Monte Carlo (MCMC) inference is then used to recover this uncertainty to a desired level of fidelity. Our approach allows the efficient evaluation of arbitrary queries over probabilistic databases with arbitrary dependencies expressed by graphical models with structure that changes during inference. MCMC sampling provides efficiency by hypothesizing modifications to possible worlds rather than generating entire worlds from scratch. Queries are then run over the portions of the world that change, avoiding the onerous cost of running full queries over each sampled world. A significant innovation of this work is the connection between MCMC sampling and materialized view maintenance techniques: we find empirically that using view maintenance techniques is several orders of magnitude faster than naively querying each sampled world. We also demonstrate our system's ability to answer relational queries with aggregation, and demonstrate additional scalability through the use of parallelization on a real-world complex model of information extraction. This framework is sufficiently expressive to support probabilistic inference not only for answering queries, but also for inferring missing database content from raw evidence.

A communication trace is a detailed record of the communication between two entities. Communication traces are vital for research in computer networks and protocols in many domains, but their release is severely constrained by privacy and security concerns. In this paper, we propose a framework in which a trace owner can match an anonymizing transformation with the requirements of analysts. The trace owner can release multiple transformed traces, each customized to an analyst’s needs, or a single transformation satisfying all requirements. The framework enables formal reasoning about anonymization policies, for example to verify that a given trace has utility for the analyst, or to obtain the most secure anonymization for the desired level of utility. Because communication traces are typically very large, we also provide techniques that allow efficient application of transformations using relational database systems. 1.

Auditing the changes to a database is critical for identifying malicious behavior, maintaining data quality, and improving system performance. But an accurate audit log is a historical record of the past that can also pose a serious threat to privacy. Policies which limit data retention conflict with the goal of accurate auditing, and data owners have to carefully balance the need for policy compliance with the goal of accurate auditing. In this paper, we provide a framework for auditing the changes to a database system while respecting data retention policies. Our framework includes a historical data model that supports flexible audit queries, along with a language for retention policies that hide individual attribute values or remove entire tuples from history. Under retention policies, the audit history is partially incomplete. We formalize the meaning of audit queries on the protected history, which can include imprecise results. We implement policy application and query answering efficiently in a standard relational system, and characterize (both theoretically and experimentally) the cases where accurate auditing can be achieved under retention restrictions.

We describe an efficient algorithm for releasing a provably private estimate of the degree distribution of a network. The algorithm satisfies a rigorous property of differential privacy, and is also extremely efficient, running on networks of 100 million nodes in a few seconds. Theoretical analysis shows that the error scales linearly with the number of unique degrees, whereas the error of conventional techniques scales linearly with the number of nodes. We complement the theoretical analysis with a thorough empirical analysis on real and synthetic graphs, showing that the algorithm's variance and bias is low, that the error diminishes as the size of the input graph increases, and that common analyses like fitting a power-law can be carried out very accurately.

We study privacy-preserving query answering over data containing relationships. A social network is a prime example of such data, where the nodes represent individuals and edges represent relationships. Nearly all interesting queries over social networks involve joins, and for such queries, existing output perturbation algorithms severely distort query answers. We propose an algorithm that significantly improves utility over competing techniques, typically reducing the error bound from polynomial in the number of nodes to polylogarithmic. The algorithm is, to the best of our knowledge, the first to answer such queries with acceptable accuracy, even for worst-case inputs.

We identify privacy risks associated with releasing network data sets and provide an algorithm that mitigates those risks. A network consists of entities connected by links representing relations such as friendship, communication, or shared activity. Maintaining privacy when publishing networked data is uniquely challenging because an individual's network context can be used to identify them even if other identifying information is removed. In this paper, we quantify the privacy risks associated with three classes of attacks on the privacy of individuals in networks, based on the knowledge used by the adversary. We show that the risks of these attacks vary greatly based on network structure and size. We propose a novel approach to anonymizing network data that models aggregate network structure and then allows samples to be drawn from that model. The approach guarantees anonymity for network entities while preserving the ability to estimate a wide variety of network measures with relatively little bias.

Auditing the changes to a database is critical for identifying malicious behavior, maintaining data quality, and improving system performance. But an accurate audit log is a historical record of the past that can also pose a serious threat to privacy. In many domains, retention policies govern how long data can be preserved by an institution. Regulations like FERPA and HIPAA (in the U.S.) or the Directive of Data Protection (in the EU), require strict retention periods to be observed, mandating the disposal of past data. In addition, institutions often adopt their own retention policies, choosing to remove sensitive data after a period of time to avoid its unintended release, or to avoid disclosure that could be forced by subpeona.

The use of any modern computer system leaves unintended traces of expired data and remnants of users' past activities. In this paper, we investigate the unintended persistence of data stored in database systems. This data can be recovered by forensic analysis, and it poses a threat to privacy.

We perform a theoretical study of the following query-view security problem: given a view V to be published, does V logically disclose information about a confidential query S? The problem is motivated by the need to manage the risk of unintended information disclosure in today's world of universal data exchange. We present a novel information-theoretic standard for query-view security. This criterion can be used to provide a precise analysis of information disclosure for a host of data exchange scenarios, including multi-party collusion and the use of outside knowledge by an adversary trying to learn privileged facts about the database. We prove a number of theoretical results for deciding security according to this standard. We also generalize our security criterion to account for prior knowledge a user or adversary may possess, and introduce techniques for measuring the magnitude of partical disclosures. We believe these results can be a foundation for practical efforts to secure data exchange frameworks, and also illuminate a nice interaction between logic and probability theory.

XPath is a language for navigating an XML document and selecting a set of element nodes. XPath expressions are used to query XML data, describe key constraints, express transformations, and reference elements in remote documents. This article studies the containment and equivalence problems for a fragment of the XPath query language, with applications in all these contexts.In particular, we study a class of XPath queries that contain branching, label wildcards and can express descendant relationships between nodes. Prior work has shown that languages that combine any two of these three features have efficient containment algorithms. However, we show that for the combination of features, containment is coNP-complete. We provide a sound and complete algorithm for containment that runs in exponential time, and study parameterized PTIME special cases. While we identify one parameterized class of queries for which containment can be decided efficiently, we also show that even with some bounded parameters, containment remains coNP-complete. In response to these negative results, we describe a sound algorithm that is efficient for all queries, but may return false negatives in some cases.

We consider the problem of evaluating a large number of XPath expressions on a stream of XML packets. We contribute two novel techniques. The first is to use a single Deterministic Finite Automaton (DFA). The contribution here is to show that the DFA can be used effectively for this problem: in our experiments we achieve a constant throughput, independently of the number of XPath expressions. The major issue is the size of the DFA, which, in theory, can be exponential in the number of XPath expressions. We provide a series of theoretical results and experimental evaluations that show that the lazy DFA has a small number of states, for all practical purposes. These results are of general interest in XPath processing, beyond stream processing. The second technique is the Streaming IndeX (SIX), which consists of adding a small amount of binary data to each XML packet that allows the query processor to achieve significant speedups. As an application of these techniques we describe the XML Toolkit (XMLTK), a collection of command-line tools providing highly scalable XML data processing.

XPath is a simple language for navigating an XML document and selecting a set of element nodes. XPath expressions are used to query XML data, describe key constraints, express transformations, and reference elements in remote documents. This paper studies the containment and equivalence problems for a fragment of the XPath query language, with applications in all these contexts.In particular, we study a class of XPath queries that contain branching, label wildcards and can express descendant relationships between nodes. Prior work has shown that languages which combine any two of these three features have efficient containment algorithms. However, we show that for the combination of features, containment is coNP-complete. We provide a sound and complete EXPTIME algorithm for containment, and study parameterized PTIME special cases. While we identify two parameterized classes of queries for which containment can be decided efficiently, we also show that even with some bounded parameters, containment is coNP-complete. In response to these negative results, we describe a sound algorithm which is efficient for all queries, but may return false negatives in some cases.