arXiv subjects

Saravanan Thirumuruganathan

Publications and source records attributed to Saravanan Thirumuruganathan.

At least 19 recordsLinked to original sources

arXiv2023

Fairness and Bias in Truth Discovery Algorithms: An Experimental Analysis

Machine learning (ML) based approaches are increasingly being used in a number of applications with societal impact. Training ML models often require vast amounts of labeled data, and crowdsourcing is a dominant paradigm for obtaining labels from multiple workers. Crowd workers may sometimes provide unreliable labels, and to address this, truth discovery (TD) algorithms such as majority voting are applied to determine the consensus labels from conflicting worker responses. However, it is important to note that these consensus labels may still be biased based on sensitive attributes such as gender, race, or political affiliation. Even when sensitive attributes are not involved, the labels can be biased due to different perspectives of subjective aspects such as toxicity. In this paper, we conduct a systematic study of the bias and fairness of TD algorithms. Our findings using two existing crowd-labeled datasets, reveal that a non-trivial proportion of workers provide biased results, and using simple approaches for TD is sub-optimal. Our study also demonstrates that popular TD algorithms are not a panacea. Additionally, we quantify the impact of these unfair workers on downstream ML tasks and show that conventional methods for achieving fairness and correcting label biases are ineffective in this setting. We end the paper with a plea for the design of novel bias-aware truth discovery algorithms that can ameliorate these issues.

cs.LG↗

arXiv2021

Fair Active Learning

Machine learning (ML) is increasingly being used in high-stakes applications impacting society. Therefore, it is of critical importance that ML models do not propagate discrimination. Collecting accurate labeled data in societal applications is challenging and costly. Active learning is a promising approach to build an accurate classifier by interactively querying an oracle within a labeling budget. We design algorithms for fair active learning that carefully selects data points to be labeled so as to balance model accuracy and fairness. We demonstrate the effectiveness and efficiency of our proposed algorithms over widely used benchmark datasets using demographic parity and equalized odds notions of fairness.

cs.LG↗

arXiv2020

Local Embeddings for Relational Data Integration

Deep learning based techniques have been recently used with promising results for data integration problems. Some methods directly use pre-trained embeddings that were trained on a large corpus such as Wikipedia. However, they may not always be an appropriate choice for enterprise datasets with custom vocabulary. Other methods adapt techniques from natural language processing to obtain embeddings for the enterprise's relational data. However, this approach blindly treats a tuple as a sentence, thus losing a large amount of contextual information present in the tuple. We propose algorithms for obtaining local embeddings that are effective for data integration tasks on relational databases. We make four major contributions. First, we describe a compact graph-based representation that allows the specification of a rich set of relationships inherent in the relational world. Second, we propose how to derive sentences from such a graph that effectively "describe" the similarity across elements (tokens, attributes, rows) in the two datasets. The embeddings are learned based on such sentences. Third, we propose effective optimization to improve the quality of the learned embeddings and the performance of integration tasks. Finally, we propose a diverse collection of criteria to evaluate relational embeddings and perform an extensive set of experiments validating them against multiple baseline methods. Our experiments show that our framework, EmbDI, produces meaningful results for data integration tasks such as schema matching and entity resolution both in supervised and unsupervised settings.

cs.DB↗

arXiv2020

Fair Active Learning

Machine learning (ML) is increasingly being used in high-stakes applications impacting society. Therefore, it is of critical importance that ML models do not propagate discrimination. Collecting accurate labeled data in societal applications is challenging and costly. Active learning is a promising approach to build an accurate classifier by interactively querying an oracle within a labeling budget. We design algorithms for fair active learning that carefully selects data points to be labeled so as to balance model accuracy and fairness. Specifically, we focus on demographic parity - a widely used measure of fairness. Extensive experiments over benchmark datasets demonstrate the effectiveness of our proposed approach.

cs.LG↗

arXiv2020

ZeroER: Entity Resolution using Zero Labeled Examples

Entity resolution (ER) refers to the problem of matching records in one or more relations that refer to the same real-world entity. While supervised machine learning (ML) approaches achieve the state-of-the-art results, they require a large amount of labeled examples that are expensive to obtain and often times infeasible. We investigate an important problem that vexes practitioners: is it possible to design an effective algorithm for ER that requires Zero labeled examples, yet can achieve performance comparable to supervised approaches? In this paper, we answer in the affirmative through our proposed approach dubbed ZeroER. Our approach is based on a simple observation -- the similarity vectors for matches should look different from that of unmatches. Operationalizing this insight requires a number of technical innovations. First, we propose a simple yet powerful generative model based on Gaussian Mixture Models for learning the match and unmatch distributions. Second, we propose an adaptive regularization technique customized for ER that ameliorates the issue of feature overfitting. Finally, we incorporate the transitivity property into the generative model in a novel way resulting in improved accuracy. On five benchmark ER datasets, we show that ZeroER greatly outperforms existing unsupervised approaches and achieves comparable performance to supervised approaches.

cs.DB↗

arXiv2019

Approximate Query Processing using Deep Generative Models

Data is generated at an unprecedented rate surpassing our ability to analyze them. The database community has pioneered many novel techniques for Approximate Query Processing (AQP) that could give approximate results in a fraction of time needed for computing exact results. In this work, we explore the usage of deep learning (DL) for answering aggregate queries specifically for interactive applications such as data exploration and visualization. We use deep generative models, an unsupervised learning based approach, to learn the data distribution faithfully such that aggregate queries could be answered approximately by generating samples from the learned model. The model is often compact - few hundred KBs - so that arbitrary AQP queries could be answered on the client side without contacting the database server. Our other contributions include identifying model bias and minimizing it through a rejection sampling based approach and an algorithm to build model ensembles for AQP for improved accuracy. Our extensive experiments show that our proposed approach can provide answers with high accuracy and low latency.

cs.DB↗

arXiv2019

DeepER -- Deep Entity Resolution

Entity resolution (ER) is a key data integration problem. Despite the efforts in 70+ years in all aspects of ER, there is still a high demand for democratizing ER - humans are heavily involved in labeling data, performing feature engineering, tuning parameters, and defining blocking functions. With the recent advances in deep learning, in particular distributed representation of words (a.k.a. word embeddings), we present a novel ER system, called DeepER, that achieves good accuracy, high efficiency, as well as ease-of-use (i.e., much less human efforts). For accuracy, we use sophisticated composition methods, namely uni- and bi-directional recurrent neural networks (RNNs) with long short term memory (LSTM) hidden units, to convert each tuple to a distributed representation (i.e., a vector), which can in turn be used to effectively capture similarities between tuples. We consider both the case where pre-trained word embeddings are available as well the case where they are not; we present ways to learn and tune the distributed representations. For efficiency, we propose a locality sensitive hashing (LSH) based blocking approach that uses distributed representations of tuples; it takes all attributes of a tuple into consideration and produces much smaller blocks, compared with traditional methods that consider only a few attributes. For ease-of-use, DeepER requires much less human labeled data and does not need feature engineering, compared with traditional machine learning based approaches which require handcrafted features, and similarity functions along with their associated thresholds. We evaluate our algorithms on multiple datasets (including benchmarks, biomedical data, as well as multi-lingual data) and the extensive experimental results show that DeepER outperforms existing solutions.

cs.DB↗

arXiv2019

Are Outlier Detection Methods Resilient to Sampling?

Outlier detection is a fundamental task in data mining and has many applications including detecting errors in databases. While there has been extensive prior work on methods for outlier detection, modern datasets often have sizes that are beyond the ability of commonly used methods to process the data within a reasonable time. To overcome this issue, outlier detection methods can be trained over samples of the full-sized dataset. However, it is not clear how a model trained on a sample compares with one trained on the entire dataset. In this paper, we introduce the notion of resilience to sampling for outlier detection methods. Orthogonal to traditional performance metrics such as precision/recall, resilience represents the extent to which the outliers detected by a method applied to samples from a sampling scheme matches those when applied to the whole dataset. We propose a novel approach for estimating the resilience to sampling of both individual outlier methods and their ensembles. We performed an extensive experimental study on synthetic and real-world datasets where we study seven diverse and representative outlier detection methods, compare results obtained from samples versus those obtained from the whole datasets and evaluate the accuracy of our resilience estimates. We observed that the methods are not equally resilient to a given sampling scheme and it is often the case that careful joint selection of both the sampling scheme and the outlier detection method is necessary. It is our hope that the paper initiates research on designing outlier detection algorithms that are resilient to sampling.

cs.LG↗

arXiv2019

Multi-Attribute Selectivity Estimation Using Deep Learning

Selectivity estimation - the problem of estimating the result size of queries - is a fundamental problem in databases. Accurate estimation of query selectivity involving multiple correlated attributes is especially challenging. Poor cardinality estimates could result in the selection of bad plans by the query optimizer. We investigate the feasibility of using deep learning based approaches for both point and range queries and propose two complementary approaches. Our first approach considers selectivity as an unsupervised deep density estimation problem. We successfully introduce techniques from neural density estimation for this purpose. The key idea is to decompose the joint distribution into a set of tractable conditional probability distributions such that they satisfy the autoregressive property. Our second approach formulates selectivity estimation as a supervised deep learning problem that predicts the selectivity of a given query. We also introduce and address a number of practical challenges arising when adapting deep learning for relational data. These include query/data featurization, incorporating query workload information in a deep learning framework and the dynamic scenario where both data and workload queries could be updated. Our extensive experiments with a special emphasis on queries with a large number of predicates and/or small result sizes demonstrates that our proposed techniques provide fast and accurate selective estimates with minimal space overhead.

cs.DB↗

arXiv2019

Data Curation with Deep Learning [Vision]

Data curation - the process of discovering, integrating, and cleaning data - is one of the oldest, hardest, yet inevitable data management problems. Despite decades of efforts from both researchers and practitioners, it is still one of the most time consuming and least enjoyable work of data scientists. In most organizations, data curation plays an important role so as to fully unlock the value of big data. Unfortunately, the current solutions are not keeping up with the ever-changing data ecosystem, because they often require substantially high human cost. Meanwhile, deep learning is making strides in achieving remarkable successes in multiple areas, such as image recognition, natural language processing, and speech recognition. In this vision paper, we explore how some of the fundamental innovations in deep learning could be leveraged to improve existing data curation solutions and to help build new ones. In particular, we provide a thorough overview of the current deep learning landscape, and identify interesting research opportunities and dispel common myths. We hope that the synthesis of these important domains will unleash a series of research activities that will lead to significantly improved solutions for many data curation tasks.

cs.DB↗

arXiv2018

Reuse and Adaptation for Entity Resolution through Transfer Learning

Entity resolution (ER) is one of the fundamental problems in data integration, where machine learning (ML) based classifiers often provide the state-of-the-art results. Considerable human effort goes into feature engineering and training data creation. In this paper, we investigate a new problem: Given a dataset D_T for ER with limited or no training data, is it possible to train a good ML classifier on D_T by reusing and adapting the training data of dataset D_S from same or related domain? Our major contributions include (1) a distributed representation based approach to encode each tuple from diverse datasets into a standard feature space; (2) identification of common scenarios where the reuse of training data can be beneficial; and (3) five algorithms for handling each of the aforementioned scenarios. We have performed comprehensive experiments on 12 datasets from 5 different domains (publications, movies, songs, restaurants, and books). Our experiments show that our algorithms provide significant benefits such as providing superior performance for a fixed training data size.

cs.DB↗

arXiv2018

Malware in the Future? Forecasting of Analyst Detection of Cyber Events

There have been extensive efforts in government, academia, and industry to anticipate, forecast, and mitigate cyber attacks. A common approach is time-series forecasting of cyber attacks based on data from network telescopes, honeypots, and automated intrusion detection/prevention systems. This research has uncovered key insights such as systematicity in cyber attacks. Here, we propose an alternate perspective of this problem by performing forecasting of attacks that are analyst-detected and -verified occurrences of malware. We call these instances of malware cyber event data. Specifically, our dataset was analyst-detected incidents from a large operational Computer Security Service Provider (CSSP) for the U.S. Department of Defense, which rarely relies only on automated systems. Our data set consists of weekly counts of cyber events over approximately seven years. Since all cyber events were validated by analysts, our dataset is unlikely to have false positives which are often endemic in other sources of data. Further, the higher-quality data could be used for a number for resource allocation, estimation of security resources, and the development of effective risk-management strategies. We used a Bayesian State Space Model for forecasting and found that events one week ahead could be predicted. To quantify bursts, we used a Markov model. Our findings of systematicity in analyst-detected cyber attacks are consistent with previous work using other sources. The advanced information provided by a forecast may help with threat awareness by providing a probable value and range for future cyber events one week ahead. Other potential applications for cyber event forecasting include proactive allocation of resources and capabilities for cyber defense (e.g., analyst staffing and sensor configuration) in CSSPs. Enhanced threat awareness may improve cybersecurity.

cs.CR↗

arXiv2018

Road Network Fusion for Incremental Map Updates

In the recent years a number of novel, automatic map-inference techniques have been proposed, which derive road-network from a cohort of GPS traces collected by a fleet of vehicles. In spite of considerable attention, these maps are imperfect in many ways: they create an abundance of spurious connections, have poor coverage, and are visually confusing. Hence, commercial and crowd-sourced mapping services heavily use human annotation to minimize the mapping errors. Consequently, their response to changes in the road network is inevitably slow. In this paper we describe \mapfuse, a system which fuses a human-annotated map (e.g., OpenStreetMap) with any automatically inferred map, thus effectively enabling quick map updates. In addition to new road creation, we study in depth road closure, which have not been examined in the past. By leveraging solid, human-annotated maps with minor corrections, we derive maps which minimize the trajectory matching errors due to both road network change and imperfect map inference of fully-automatic approaches.

cs.OH↗

arXiv2017

A Cost-based Optimizer for Gradient Descent Optimization

As the use of machine learning (ML) permeates into diverse application domains, there is an urgent need to support a declarative framework for ML. Ideally, a user will specify an ML task in a high-level and easy-to-use language and the framework will invoke the appropriate algorithms and system configurations to execute it. An important observation towards designing such a framework is that many ML tasks can be expressed as mathematical optimization problems, which take a specific form. Furthermore, these optimization problems can be efficiently solved using variations of the gradient descent (GD) algorithm. Thus, to decouple a user specification of an ML task from its execution, a key component is a GD optimizer. We propose a cost-based GD optimizer that selects the best GD plan for a given ML task. To build our optimizer, we introduce a set of abstract operators for expressing GD algorithms and propose a novel approach to estimate the number of iterations a GD algorithm requires to converge. Extensive experiments on real and synthetic datasets show that our optimizer not only chooses the best GD plan but also allows for optimizations that achieve orders of magnitude performance speed-up.

cs.DB↗

arXiv2017

Kharita: Robust Map Inference using Graph Spanners

The widespread availability of GPS information in everyday devices such as cars, smartphones and smart watches make it possible to collect large amount of geospatial trajectory information. A particularly important, yet technically challenging, application of this data is to identify the underlying road network and keep it updated under various changes. In this paper, we propose efficient algorithms that can generate accurate maps in both batch and online settings. Our algorithms utilize techniques from graph spanners so that they produce maps can effectively handle a wide variety of road and intersection shapes. We conduct a rigorous evaluation of our algorithms over two real-world datasets and under a wide variety of performance metrics. Our experiments show a significant improvement over prior work. In particular, we observe an increase in Biagioni f-score of up to 20% when compared to the state of the art while reducing the execution time by an order of magnitude. We also make our source code open source for reproducibility and enable other researchers to build on our work.

cs.OH↗

arXiv2016

Discovering the Skyline of Web Databases

Many web databases are "hidden" behind proprietary search interfaces that enforce the top-$k$ output constraint, i.e., each query returns at most $k$ of all matching tuples, preferentially selected and returned according to a proprietary ranking function. In this paper, we initiate research into the novel problem of skyline discovery over top-$k$ hidden web databases. Since skyline tuples provide critical insights into the database and include the top-ranked tuple for every possible ranking function following the monotonic order of attribute values, skyline discovery from a hidden web database can enable a wide variety of innovative third-party applications over one or multiple web databases. Our research in the paper shows that the critical factor affecting the cost of skyline discovery is the type of search interface controls provided by the website. As such, we develop efficient algorithms for three most popular types, i.e., one-ended range, free range and point predicates, and then combine them to support web databases that feature a mixture of these types. Rigorous theoretical analysis and extensive real-world online and offline experiments demonstrate the effectiveness of our proposed techniques and their superiority over baseline solutions.

cs.DB↗

arXiv2016

HDBSCAN: Density based Clustering over Location Based Services

Location Based Services (LBS) have become extremely popular and used by millions of users. Popular LBS run the entire gamut from mapping services (such as Google Maps) to restaurants (such as Yelp) and real-estate (such as Redfin). The public query interfaces of LBS can be abstractly modeled as a kNN interface over a database of two dimensional points: given an arbitrary query point, the system returns the k points in the database that are nearest to the query point. Often, k is set to a small value such as 20 or 50. In this paper, we consider the novel problem of enabling density based clustering over an LBS with only a limited, kNN query interface. Due to the query rate limits imposed by LBS, even retrieving every tuple once is infeasible. Hence, we seek to construct a cluster assignment function f(.) by issuing a small number of kNN queries, such that for any given tuple t in the database which may or may not have been accessed, f(.) outputs the cluster assignment of t with high accuracy. We conduct a comprehensive set of experiments over benchmark datasets and popular real-world LBS such as Yahoo! Flickr, Zillow, Redfin and Google Maps.

cs.DB↗

arXiv2015

Aggregate Estimations over Location Based Services

Location based services (LBS) have become very popular in recent years. They range from map services (e.g., Google Maps) that store geographic locations of points of interests, to online social networks (e.g., WeChat, Sina Weibo, FourSquare) that leverage user geographic locations to enable various recommendation functions. The public query interfaces of these services may be abstractly modeled as a kNN interface over a database of two dimensional points on a plane: given an arbitrary query point, the system returns the k points in the database that are nearest to the query point. In this paper we consider the problem of obtaining approximate estimates of SUM and COUNT aggregates by only querying such databases via their restrictive public interfaces. We distinguish between interfaces that return location information of the returned tuples (e.g., Google Maps), and interfaces that do not return location information (e.g., Sina Weibo). For both types of interfaces, we develop aggregate estimation algorithms that are based on novel techniques for precisely computing or approximately estimating the Voronoi cell of tuples. We discuss a comprehensive set of real-world experiments for testing our algorithms, including experiments on Google Maps, WeChat, and Sina Weibo.

cs.DB↗