Many modern machine learning (ML) methods rely on embedding models to learn vector representations (embeddings) for a set of entities (embedding tables). As increasingly diverse ML applications utilize embedding models and embedding tables continue to grow in size and number, there has been a surge in the ad-hoc development of specialized frameworks targeted to train large embedding models for specific tasks. Although the scalability issues that arise in different embedding model training tasks are similar, each of these frameworks independently reinvents and customizes storage components for specific tasks, leading to substantial duplicated engineering efforts in both development and deployment. This paper presents MLKV, an efficient, extensible, and reusable data storage framework designed to address the scalability challenges in embedding model training, specifically data stall and staleness. MLKV augments disk-based key-value storage by democratizing optimizations that were previously exclusive to individual specialized frameworks and provides easy-to-use interfaces for embedding model training tasks. Extensive experiments on open-source workloads, as well as applications in eBay's payment transaction risk detection and seller payment risk detection, show that MLKV outperforms offloading strategies built on top of industrial-strength key-value stores by 1.6-12.6 ×. MLKV is open-source at https://github.com/llm-db/MLKV.

Abstract There have been many decades of work on optimizing query processing in database management systems. Recently, modern machine learning (ML), and specifically reinforcement learning (RL), have gained increased attention as a means to develop a query optimizer (QO). In this work, we take a closer look at two recent RL-based QO methods to better understand their behavior. We find that these RL-based methods do not generalize as well as it seems at first glance. Thus, we ask a simple question: How do RL-based QOs compare to a simple, modern, adaptive query processing approach? To answer this question, we chose two simple adaptive query processing techniques and implemented them in PostgreSQL. The first adapts an individual join operation on-the-fly and switches between a Nested Loop Join algorithm and a Hash Join algorithm to avoid sub-optimal join algorithm decisions. The second is a technique called Lookahead Information Passing (LIP), in which adaptive semijoin techniques are used to make a pipeline of join operations execute efficiently. To our surprise, we find that this simple adaptive query processing approach is not only competitive to these RL-based approaches but, in some cases, outperforms the RL-based approaches. The adaptive approach is also appealing because it does not require an expensive training step, and it is fully interpretable compared to the RL-based QO approaches. Further, the adaptive method works across complex query constructs that RL-based QO methods currently cannot optimize.

Datasets that pair Knowledge Graphs (KG) and text together (KG-T) can be used to train forward and reverse neural models that generate text from KG and vice versa. However models trained on datasets where KG and text pairs are not equivalent can suffer from more hallucination and poorer recall. In this paper, we verify this empirically by generating datasets with different levels of noise and find that noisier datasets do indeed lead to more hallucination. We argue that the ability of forward and reverse models trained on a dataset to cyclically regenerate source KG or text is a proxy for the equivalence between the KG and the text in the dataset. Using cyclic evaluation we find that manually created WebNLG is much better than automatically created TeKGen and T-REx. Informed by these observations, we construct a new, improved dataset called LAGRANGE using heuristics meant to improve equivalence between KG and text and show the impact of each of the heuristics on cyclic evaluation. We also construct two synthetic datasets using large language models (LLMs), and observe that these are conducive to models that perform significantly well on cyclic generation of text, but less so on cyclic generation of KGs, probably because of a lack of a consistent underlying ontology.

Methods for carefully selecting or generating a small set of training data to learn from, i.e., data pruning, coreset selection, and dataset distillation, have been shown to be effective in reducing the ever-increasing cost of training neural networks. Behind this success are rigorously designed, yet expensive, strategies for identifying the most informative training examples out of large datasets. In this work, we revisit these methods to understand if the additional computational costs associated with such strategies are justified from the perspective of time-to-accuracy, which has become a critical efficiency measure of deep neural network training over large datasets. Surprisingly, we find that many of the recently proposed methods underperform what we call Repeated Sampling of Random Subsets (RSRS or RS2), a powerful yet overlooked extension of the standard random baseline that learns from repeatedly sampled data throughout training instead of a fixed random subset. We test RS2 against thirty-two state-of-the-art data pruning and distillation methods across four datasets including ImageNet. Our results demonstrate that RS2 significantly reduces time-to-accuracy, particularly in practical regimes where accuracy, but not runtime, is similar to that of training on full dataset. For example, when training ResNet-18 on ImageNet, with 10\% of the dataset each epoch RS2 reaches an accuracy of 66\% versus 69\% when training with the full dataset. The best competing method achieves only 55\% while training 1.6$\times$ slower than RS2. Beyond the above meta-study, we discuss the theoretical properties of RS2 such as its convergence rate and generalization error. Our primary goal is to highlight that future works that aim to minimize total training cost by using subset selection, need to consider 1) the total computation cost (including preparing the subset) and 2) should aim to outperform a simple extension of random sampling (i.e., RS2).

Finetuning foundation models for specific tasks is an emerging paradigm in modern machine learning. The efficacy of task-specific finetuning largely depends on the selection of appropriate training data. We present TSDS (Task-Specific Data Selection), a framework to select data for task-specific model finetuning, guided by a small but representative set of examples from the target task. To do so, we formulate data selection for task-specific finetuning as an optimization problem with a distribution alignment loss based on optimal transport to capture the discrepancy between the selected data and the target distribution. In addition, we add a regularizer to encourage the diversity of the selected data and incorporate kernel density estimation into the regularizer to reduce the negative effects of near-duplicates among the candidate data. We connect our optimization problem to nearest neighbor search and design efficient algorithms to compute the optimal solution based on approximate nearest neighbor search techniques. We evaluate our method on data selection for both continued pretraining and instruction tuning of language models. We show that instruction tuning using data selected by our method with a 1\% selection ratio often outperforms using the full dataset and beats the baseline selection methods by 1.5 points in F1 score on average.

We study training of Graph Neural Networks (GNNs) for large-scale graphs. We revisit the premise of using distributed training for billion-scale graphs and show that for graphs that fit in main memory or the SSD of a single machine, out-of-core pipelined training with a single GPU can outperform state-of-the-art (SoTA) multi-GPU solutions. We introduce MariusGNN, the first system that utilizes the entire storage hierarchy---including disk---for GNN training. MariusGNN introduces a series of data organization and algorithmic contributions that 1) minimize the end-to-end time required for training and 2) ensure that models learned with disk-based training exhibit accuracy similar to those fully trained in memory. We evaluate MariusGNN against SoTA systems for learning GNN models and find that single-GPU training in MariusGNN achieves the same level of accuracy up to 8× faster than multi-GPU training in these systems, thus, introducing an order of magnitude monetary cost reduction. MariusGNN is open-sourced at www.marius-project.org.

Vector search has emerged as the foundation for large-scale information retrieval and machine learning systems, with search engines like Google and Bing processing tens of thousands of queries per second on petabyte-scale document datasets by evaluating vector similarities between encoded query texts and web documents. As performance demands for vector search systems surge, accelerated hardware offers a promising solution in the post-Moore's Law era. We introduce FANNS, an end-to-end and scalable vector search framework on FPGAs. Given a user-provided recall requirement on a dataset and a hardware resource budget, FANNS automatically co-designs hardware and algorithm, subsequently generating the corresponding accelerator. The framework also supports scale-out by incorporating a hardware TCP/IP stack in the accelerator. FANNS attains up to 23.0× and 37.2× speedup compared to FPGA and CPU baselines, respectively, and demonstrates superior scalability to GPUs, achieving 5.5× and 7.6× speedup in median and 95th percentile (P95) latency within an eight-accelerator configuration. The remarkable performance of FANNS lays a robust groundwork for future FPGA integration in data centers and AI supercomputers.

There is an increasing adoption of machine learning for encoding data into vectors to serve online recommendation and search use cases. As a result, recent data management systems propose augmenting query processing with online vector similarity search. In this work, we explore vector similarity search in the context of Knowledge Graphs (KGs). Motivated by the tasks of finding related KG queries and entities for past KG query workloads, we focus on hybrid vector similarity search (hybrid queries for short) where part of the query corresponds to vector similarity search and part of the query corresponds to predicates over relational attributes associated with the underlying data vectors. For example, given past KG queries for a song entity, we want to construct new queries for new song entities whose vector representations are close to the vector representation of the entity in the past KG query. But entities in a KG also have non-vector attributes such as a song associated with an artist, a genre, and a release date. Therefore, suggested entities must also satisfy query predicates over non-vector attributes beyond a vector-based similarity predicate. While these tasks are central to KGs, our contributions are generally applicable to hybrid queries. In contrast to prior works that optimize online queries, we focus on enabling efficient batch processing of past hybrid query workloads. We present our system, HQI, for high-throughput batch processing of hybrid queries. We introduce a workload-aware vector data partitioning scheme to tailor the vector index layout to the given workload and describe a multi-query optimization technique to reduce the overhead of vector similarity computations. We evaluate our methods on industrial workloads and demonstrate that HQI yields a 31× improvement in throughput for finding related KG queries compared to existing hybrid query processing approaches.

There have been many decades of work on optimizing query processing in database management systems. Recently, modern machine learning (ML), and specifically reinforcement learning (RL), has gained increased attention as a means to develop a query optimizer (QO). In this work, we take a closer look at two recent state-of-the-art (SOTA) RL-based QO methods to better understand their behavior. We find that these RL-based methods do not generalize as well as it seems at first glance. Thus, we ask a simple question: How do SOTA RL-based QOs compare to a simple, modern, adaptive query processing approach? To answer this question, we choose two simple adaptive query processing techniques and implemented them in PostgreSQL. The first adapts an individual join operation on-the-fly and switches between a Nested Loop Join algorithm and a Hash Join algorithm to avoid sub-optimal join algorithm decisions. The second is a technique called Lookahead Information Passing (LIP), in which adaptive semijoin techniques are used to make a pipeline of join operations execute efficiently. To our surprise, we find that this simple adaptive query processing approach is not only competitive to the SOTA RL-based approaches but, in some cases, outperforms the RL-based approaches. The adaptive approach is also appealing because it does not require an expensive training step, and it is fully interpretable compared to the RL-based QO approaches. Further, the adaptive method works across complex query constructs that RL-based QO methods currently cannot optimize.

We introduce Saga, a next-generation knowledge construction and serving platform for powering knowledge-based applications at industrial scale. Saga follows a hybrid batch-incremental design to continuously integrate billions of facts about real-world entities and construct a central knowledge graph that supports multiple production use cases with diverse requirements around data freshness, accuracy, and availability. In this paper, we discuss the unique challenges associated with knowledge graph construction at industrial scale, and review the main components of Saga and how they address these challenges. Finally, we share lessons-learned from a wide array of production use cases powered by Saga.

We study the problem of robust mean estimation and introduce a novel Hamming distance-based measure of distribution shift for coordinate-level corruptions. We show that this measure yields adversary models that capture more realistic corruptions than those used in prior works, and present an information-theoretic analysis of robust mean estimation in these settings. We show that for structured distributions, methods that leverage the structure yield information theoretically more accurate mean estimation. We also focus on practical algorithms for robust mean estimation and study when data cleaning-inspired approaches that first fix corruptions in the input data and then perform robust mean estimation can match the information theoretic bounds of our analysis. We finally demonstrate experimentally that this two-step approach outperforms structure-agnostic robust estimation and provides accurate mean estimation even for high-magnitude corruption.

Graph embeddings have emerged as the de facto representation for modern machine learning over graph data structures. The goal of graph embedding models is to convert high-dimensional sparse graphs into low-dimensional, dense and continuous vector spaces that preserve the graph structure properties. However, learning a graph embedding model is a resource intensive process, and existing solutions rely on expensive distributed computation to scale training to instances that do not fit in GPU memory. This demonstration showcases Marius: a new open-source engine for learning graph embedding models over billion-edge graphs on a single machine. Marius is built around a recently-introduced architecture for machine learning over graphs that utilizes pipelining and a novel data replacement policy to maximize GPU utilization and exploit the entire memory hierarchy (including disk, CPU, and GPU memory) to scale to large instances. The audience will experience how to develop, train, and deploy graph embedding models using Marius' configuration-driven programming model. Moreover, the audience will have the opportunity to explore Marius' deployments on applications including link-prediction on WikiKG90M and reasoning queries on a paleobiology knowledge graph. Marius is available as open source software at https://marius-project.org.

Structured data, or data that adheres to a pre-defined schema, can suffer from fragmented context: information describing a single entity can be scattered across multiple datasets or tables tailored for specific business needs, with no explicit linking keys. Context enrichment, or rebuilding fragmented context, using keyless joins is an implicit or explicit step in machine learning (ML) pipelines over structured data sources. This process is tedious, domain-specific, and lacks support in now-prevalent no-code ML systems that let users create ML pipelines using just input data and high-level configuration files. In response, we propose Ember, a system that abstracts and automates keyless joins to generalize context enrichment. Our key insight is that Ember can enable a general keyless join operator by constructing an index populated with task-specific embeddings. Ember learns these embeddings by leveraging Transformer-based representation learning techniques. We describe our architectural principles and operators when developing Ember, and empirically demonstrate that Ember allows users to develop no-code context enrichment pipelines for five domains, including search, recommendation and question answering, and can exceed alternatives by up to 39% recall, with as little as a single line configuration change.

Directed graphical models (DGMs) are a class of probabilistic models that are widely used for predictive analysis in sensitive domains such as medical diagnostics. In this paper, we present an algorithm for differentially-private learning of the parameters of a DGM. Our solution optimizes for the utility of inference queries over the DGM and \emph{adds noise that is customized to the properties of the private input dataset and the graph structure of the DGM}. To the best of our knowledge, this is the first explicit data-dependent privacy budget allocation algorithm in the context of DGMs. We compare our algorithm with a standard data-independent approach over a diverse suite of benchmarks and demonstrate that our solution requires a privacy budget that is roughly $3\times$ smaller to obtain the same or higher utility.

We study the problem of discovering functional dependencies (FD) from a noisy data set. We adopt a statistical perspective and draw connections between FD discovery and structure learning in probabilistic graphical models. We show that discovering FDs from a noisy data set is equivalent to learning the structure of a model over binary random variables, where each random variable corresponds to a functional of the data set attributes. We build upon this observation to introduce FDX a conceptually simple framework in which learning functional dependencies corresponds to solving a sparse regression problem. We show that FDX can recover true functional dependencies across a diverse array of real-world and synthetic data sets, even in the presence of noisy or missing data. We find that FDX scales to large data instances with millions of tuples and hundreds of attributes while it yields an average F1 improvement of 2x against state-of-the-art FD discovery methods.

Crowdsourcing is essential for collecting information about real-world entities. Existing crowdsourced data extraction solutions use fixed, non-adaptive querying strategies that repeatedly ask workers to provide entities from a fixed domain until a desired level of coverage is reached. Unfortunately, such solutions are highly impractical as they yield many duplicate extractions. We design an adaptive querying framework, CRUX, that maximizes the number of extracted entities for a given budget. We show that the problem of budgeted crowdsourced entity extraction is NP-Hard. We leverage two insights to focus our extraction efforts: \em exploiting the structure of the domain of interest, and \em using exclude lists to limit repeated extractions. We develop new statistical tools to reason about the number of new distinct extracted entities of \em additional queries under the presence of little information, and embed them within adaptive algorithms that maximize the distinct extracted entities under budget constraints. We evaluate our techniques on synthetic and real-world datasets, demonstrating an improvement of up to 300% over competing approaches for the same budget.

Most theoretical frameworks that focus on data errors and inconsistencies follow logic-based reasoning. Yet, practical data cleaning tools need to incorporate statistical reasoning to be effective in real-world data cleaning tasks. Motivated by empirical successes, we propose a formal framework for unclean databases, where two types of statistical knowledge are incorporated: The first represents a belief of how intended (clean) data is generated, and the second represents a belief of how noise is introduced in the actual observed database. To capture this noisy channel model, we introduce the concept of a Probabilistic Unclean Database (PUD), a triple that consists of a probabilistic database that we call the intention, a probabilistic data transformator that we call the realization and captures how noise is introduced, and an observed unclean database that we call the observation. We define three computational problems in the PUD framework: cleaning (infer the most probable intended database, given a PUD), probabilistic query answering (compute the probability of an answer tuple over the unclean observed database), and learning (estimate the most likely intention and realization models of a PUD, given examples as training data). We illustrate the PUD framework on concrete representations of the intention and realization, show that they generalize traditional concepts of repairs such as cardinality and value repairs, draw connections to consistent query answering, and prove tractability results. We further show that parameters can be learned in some practical instantiations, and in fact, prove that under certain conditions we can learn a PUD directly from a single dirty database without any need for clean examples.

As data volume and variety have increased, so have the ties between machine learning and data integration become stronger. For machine learning to be effective, one must utilize data from the greatest possible variety of sources; and this is why data integration plays a key role. At the same time machine learning is driving automation in data integration, resulting in overall reduction of integration costs and improved accuracy. This tutorial focuses on three aspects of the synergistic relationship between data integration and machine learning: (1) we survey how state-of-the-art data integration solutions rely on machine learning-based approaches for accurate results and effective human-in-the-loop pipelines, (2) we review how end-to-end machine learning applications rely on data integration to identify accurate, clean, and relevant data for their analytics exercises, and (3) we discuss open research challenges and opportunities that span across data integration and machine learning.

We introduce a few-shot learning framework for error detection. We show that data augmentation (a form of weak supervision) is key to training high-quality, ML-based error detection models that require minimal human involvement. Our framework consists of two parts: (1) an expressive model to learn rich representations that capture the inherent syntactic and semantic heterogeneity of errors; and (2) a data augmentation model that, given a small seed of clean records, uses dataset-specific transformations to automatically generate additional training data. Our key insight is to learn data augmentation policies from the noisy input dataset in a weakly supervised manner. We show that our framework detects errors with an average precision of ~94% and an average recall of ~93% across a diverse array of datasets that exhibit different types and amounts of errors. We compare our approach to a comprehensive collection of error detection methods, ranging from traditional rule-based methods to ensemble-based and active learning approaches. We show that data augmentation yields an average improvement of 20 F1 points while it requires access to 3x fewer labeled examples compared to other ML approaches.

AI/ML is becoming a horizontal technology: its application is expanding to more domains, and its integration touches more parts of the technology stack. Given the strong dependence of ML on data, this expansion creates a new space for applying data management techniques. At the same time, the deeper integration of ML in the technology stack provides more touch points where ML can be used in data management systems and vice versa. In this panel, we invite researchers working in this domain to discuss this emerging world and its implications on data-management research. Among other topics, the discussion will touch on the opportunities for interesting research, how we can interact with other communities, what is the core expertise we bring to the table, and how we can conduct and evaluate this research effectively within our own community. The goal of the panel is to nudge the community to appreciate the opportunities in this new world of horizontal AI/ML and to spur a discussion on how we can shape an effective research agenda.

We focus on knowledge base construction (KBC) from richly formatted data. In contrast to KBC from text or tabular data, KBC from richly formatted data aims to extract relations conveyed jointly via textual, structural, tabular, and visual expressions. We introduce Fonduer, a machine-learning-based KBC system for richly formatted data. Fonduer presents a new data model that accounts for three challenging characteristics of richly formatted data: (1) prevalent document-level relations, (2) multimodality, and (3) data variety. Fonduer uses a new deep-learning model to automatically capture the representation (i.e., features) needed to learn how to extract relations from richly formatted data. Finally, Fonduer provides a new programming model that enables users to convert domain expertise, based on multiple modalities of information, to meaningful signals of supervision for training a KBC system. Fonduer-based KBC systems are in production for a range of use cases, including at a major online retailer. We compare Fonduer against state-of-the-art KBC approaches in four different domains. We show that Fonduer achieves an average improvement of 41 F1 points on the quality of the output knowledge base-and in some cases produces up to 1.87× the number of correct entries-compared to expert-curated public knowledge bases. We also conduct a user study to assess the usability of Fonduer's new programming model. We show that after using Fonduer for only 30 minutes, non-domain experts are able to design KBC systems that achieve on average 23 F1 points higher quality than traditional machine-learning-based KBC approaches.

There is now more data to analyze than ever before. As data volume and variety have increased, so have the ties between machine learning and data integration become stronger. For machine learning to be effective, one must utilize data from the greatest possible variety of sources; and this is why data integration plays a key role. At the same time machine learning is driving automation in data integration, resulting in overall reduction of integration costs and improved accuracy. This tutorial focuses on three aspects of the synergistic relationship between data integration and machine learning: (1) we survey how state-of-the-art data integration solutions rely on machine learning-based approaches for accurate results and effective human-in-the-loop pipelines, (2) we review how end-to-end machine learning applications rely on data integration to identify accurate, clean, and relevant data for their analytics exercises, and (3) we discuss open research challenges and opportunities that span across data integration and machine learning.

Entity matching (EM) finds data instances that refer to the same real-world entity. In this paper we examine applying deep learning (DL) to EM, to understand DL's benefits and limitations. We review many DL solutions that have been developed for related matching tasks in text processing (e.g., entity linking, textual entailment, etc.). We categorize these solutions and define a space of DL solutions for EM, as embodied by four solutions with varying representational power: SIF, RNN, Attention, and Hybrid. Next, we investigate the types of EM problems for which DL can be helpful. We consider three such problem types, which match structured data instances, textual instances, and dirty instances, respectively. We empirically compare the above four DL solutions with Magellan, a state-of-the-art learning-based EM solution. The results show that DL does not outperform current solutions on structured EM, but it can significantly outperform them on textual and dirty EM. For practitioners, this suggests that they should seriously consider using DL for textual and dirty EM problems. Finally, we analyze DL's performance and discuss future research directions.

As data volume and variety have increased, so have the ties between machine learning and data integration become stronger. For machine learning to be effective, one must utilize data from the greatest possible variety of sources; and this is why data integration plays a key role. At the same time machine learning is driving automation in data integration, resulting in overall reduction of integration costs and improved accuracy. This tutorial focuses on three aspects of the synergistic relationship between data integration and machine learning: (1) we survey how state-of-the-art data integration solutions rely on machine learning-based approaches for accurate results and effective human-in-the-loop pipelines, (2) we review how end-to-end machine learning applications rely on data integration to identify accurate, clean, and relevant data for their analytics exercises, and (3) we discuss open research challenges and opportunities that span across data integration and machine learning.

We focus on data fusion, i.e., the problem of unifying conflicting data from data sources into a single representation by estimating the source accuracies. We propose SLiMFast, a framework that expresses data fusion as a statistical learning problem over discriminative probabilistic models, which in many cases correspond to logistic regression. In contrast to previous approaches that use complex generative models, discriminative models make fewer distributional assumptions over data sources and allow us to obtain rigorous theoretical guarantees. Furthermore, we show how SLiMFast enables incorporating domain knowledge into data fusion, yielding accuracy improvements of up to 50% over state-of-the-art baselines. Building upon our theoretical results, we design an optimizer that obviates the need for users to manually select an algorithm for learning SLiMFast's parameters. We validate our optimizer on multiple real-world datasets and show that it can accurately predict the learning algorithm that yields the best data fusion results.

We introduce HoloClean, a framework for holistic data repairing driven by probabilistic inference. HoloClean unifies qualitative data repairing, which relies on integrity constraints or external data sources, with quantitative data repairing methods, which leverage statistical properties of the input data. Given an inconsistent dataset as input, HoloClean automatically generates a probabilistic program that performs data repairing. Inspired by recent theoretical advances in probabilistic inference, we introduce a series of optimizations which ensure that inference over HoloClean's probabilistic model scales to instances with millions of tuples. We show that HoloClean finds data repairs with an average precision of ∼ 90% and an average recall of above ∼ 76% across a diverse array of datasets exhibiting different types of errors. This yields an average F1 improvement of more than 2× against state-of-the-art methods.

Recently there has been a rapid increase in the number of data sources and data services, such as cloud-based data markets and data portals, that facilitate the collection, publishing and trading of data. Data sources typically exhibit large heterogeneity in the type and quality of data they provide. Unfortunately, when the number of data sources is large, it is difficult for users to reason about the actual usefulness of sources for their applications and the trade-offs between the benefits and costs of acquiring and integrating sources. In this demonstration we present \textsc{SourceSight}, a system that allows users to interactively explore a large number of heterogeneous data sources, and discover valuable sets of sources for diverse integration tasks. \textsc{SourceSight}~uses a novel multi-level source quality index that enables effective source selection at different granularity levels, and introduces a collection of new techniques to discover and evaluate relevant sources for integration.

Understanding the diffusion of information in social network and social media requires modeling the text diffusion process. In this work, we develop the HawkesTopic model (HTM) for analyzing text-based cascades, such as "retweeting a post" or "publishing a follow-up blog post". HTM combines Hawkes processes and topic modeling to simultaneously reason about the information diffusion pathways and the topics characterizing the observed textual information. We show how to jointly infer them with a mean-field variational inference algorithm and validate our approach on both synthetic and real-world data sets, including a news media dataset for modeling information diffusion, and an ArXiv publication dataset for modeling scientific influence. The results show that HTM is significantly more accurate than several baselines for both tasks.

Rapidly increasing volumes of news feeds from diverse data sources, such as online newspapers, Twitter and online blogs are proving to be extremely valuable resources in helping anticipate, detect, and forecast outbreaks of rare diseases. This paper presents SourceSeer, a novel algorithmic framework that combines spatio-temporal topic models with sourcebased anomaly detection techniques to effectively forecast the emergence and progression of infectious rare diseases. SourceSeer is capable of discovering the location focus of each source allowing sources to be used as experts with varying degrees of authoritativeness. To fuse the individual source predictions into a final outbreak prediction we employ a multiplicative weights algorithm taking into account the accuracy of each source. We evaluate the performance of SourceSeer using incidence data for hantavirus syndromes in multiple countries of Latin America provided by HealthMap over a timespan of fifteen months. We demonstrate that SourceSeer makes predictions of increased accuracy compared to several baselines and is capable of forecasting disease outbreaks in a timely manner even when no outbreaks were previously reported.

As the world evolves around us, so does the digital coverage of it. Events of diverse types, associated with different actors and various locations, are continuously captured by multiple information sources such as news articles, blogs, social media etc. day by day. In the digital world, these events are represented through information snippets that contain information on the involved entities, a description of the event, when the event occurred, etc. In our work, we observe that events (and their corresponding digital representations) are often inter-connected, i.e., they form stories which represent evolving relationships between events over time. Take as an example the plane crash in Ukraine in July 2014 which involved multiple entities such as "Ukraine", "Malaysia", and "Russia" and multiple events ranging from the actual crash to the incident investigation and the presentation of the investigator's findings. In this demonstration we present StoryPivot, a framework that helps its users to detect evolving stories in event datasets over time. To resolve stories, we differentiate between story identification, the problem of connecting events over time within a source, and story alignment, the problem of integrating stories across sources. The goal of this demonstration is to present an interactive exploration of both these problems and how events can be dynamically interpreted and put into context in real-world datasets.

Data integration is a challenging task due to the large numbers of autonomous data sources. This necessitates the development of techniques to reason about the benefits and costs of acquiring and integrating data. Recently the problem of source selection (i.e., identifying the subset of sources that maximizes the profit from integration) was introduced as a preprocessing step before the actual integration. The problem was studied for static sources and used the accuracy of data fusion to quantify the integration profit.

We present SPARSI, a novel theoretical framework for partitioning sensitive data across multiple non-colluding adversaries. Most work in privacy-aware data sharing has considered disclosing summaries where the aggregate information about the data is preserved, but sensitive user information is protected. Nonetheless, there are applications, including online advertising, cloud computing and crowdsourcing markets, where detailed and fine-grained user data must be disclosed. We consider a new data sharing paradigm and introduce the problem of privacy-aware data partitioning, where a sensitive dataset must be partitioned among k untrusted parties (adversaries). The goal is to maximize the utility derived by partitioning and distributing the dataset, while minimizing the total amount of sensitive information disclosed. The data should be distributed so that an adversary, without colluding with other adversaries, cannot draw additional inferences about the private information, by linking together multiple pieces of information released to her. The assumption of no collusion is both reasonable and necessary in the above application domains that require release of private user information. SPARSI enables us to formally define privacy-aware data partitioning using the notion of sensitive properties for modeling private information and a hypergraph representation for describing the interdependencies between data entries and private information. We show that solving privacy-aware partitioning is, in general, NP-hard, but for specific information disclosure functions, good approximate solutions can be found using relaxation techniques. Finally, we present a local search algorithm applicable to generic information disclosure functions. We conduct a rigorous performance evaluation with real-world and synthetic datasets that illustrates the effectiveness of SPARSI at partitioning sensitive data while minimizing disclosure.

While extensive work has been done on evaluating queries over tuple-independent probabilistic databases, query evaluation over correlated data has received much less attention even though the support for correlations is essential for many natural applications of probabilistic databases, e.g., information extraction, data integration, computer vision, etc. In this paper, we develop a novel approach for efficiently evaluating probabilistic queries over correlated databases where correlations are represented using a factor graph, a class of graphical models widely used for capturing correlations and performing statistical inference. Our approach exploits the specific values of the factor parameters and the determinism in the correlations, collectively called local structure, to reduce the complexity of query evaluation. Our framework is based on arithmetic circuits, factorized representations of probability distributions that can exploit such local structure. Traditionally, arithmetic circuits are generated following a compilation process and can not be updated directly. We introduce a generalization of arithmetic circuits, called annotated arithmetic circuits, and a novel algorithm for updating them, which enables us to answer probabilistic queries efficiently. We present a comprehensive experimental analysis and show speed-ups of at least one order of magnitude in many cases.