arXiv subjects

Aécio Santos

Publications and source records attributed to Aécio Santos.

12 recordsLinked to original sources

arXiv2026

The Silent Failures in Agentic Data Science

Large language models have significantly lowered the barrier to writing code. Naturally, this has also led to the development of data science agents that promise automating the design and execution of data science from task formulation to model development. Existing evaluations of agentic data science (agentic DS) assess downstream predictive performance and code correctness, providing preliminary insight in agentic DS capabilities. We argue that these evaluations overlook the fundamental principles of data science: methodological data science- and domain knowledge. As such, only assessing downstream performance masks methodological failures such as data leakage and inappropriate validation, limiting the development of robust agentic DS systems. In this paper, we evaluate agentic DS beyond accuracy from three different angles: 1) a large-scale analysis of failure modes in common agentic DS frameworks and models across 220 data science tasks, 2) a reusable and extensible test suite with curated data science tasks that test for six specific failures in a controlled manner, and 3) a qualitative assessment of agentic DS pipelines against expert solutions on Kaggle competitions. We find that the majority of agentic solutions contain at least one failure mode despite the complexity of agentic harness or model capacity. Moreover, we identify a positive correlation between the occurrence of data leakage failures and predictive performance, confirming the necessity to evaluate agents beyond accuracy. Our qualitative analysis shows that DS agents generally lack problem solving creativity and domain knowledge, compared to competition-winning data scientists. We believe that principled failure mode analysis, as presented here, should be integral part of agentic DS evaluations, and that the insights we surface will be key in developing trustworthy agentic DS pipelines.

cs.AI↗

arXiv2025

AutoDDG: Automated Dataset Description Generation using Large Language Models

The proliferation of datasets across open data portals and enterprise data lakes presents an opportunity for deriving data-driven insights. Widely-used dataset search systems rely on keyword search over dataset metadata, including descriptions, to support discovery. Therefore, when these descriptions are incomplete, missing, or inconsistent with dataset contents, findability is severely compromised. To improve findability, we introduce AutoDDG, a framework that automatically generates descriptions of tabular data. By adopting a data-driven approach to summarize dataset contents and leveraging large language models (LLMs) to enrich summaries with semantic information and produce human-readable text, AutoDDG derives descriptions that are comprehensive, accurate, readable, and concise. A critical challenge in this problem is evaluating the effectiveness of description generation methods and assessing the quality of the generated descriptions. We propose a comprehensive evaluation methodology that combines retrieval, reference-based, and reference-free assessment, with human validation. Our experimental results using new benchmarks demonstrate that AutoDDG generates high-quality, accurate descriptions at scale, significantly improving dataset retrieval performance across diverse use cases. AutoDDG is publicly available at https://github.com/VIDA-NYU/AutoDDG.

cs.DB↗

arXiv2025

Interactive Data Harmonization with LLM Agents: Opportunities and Challenges

Data harmonization is an essential task that entails integrating datasets from diverse sources. Despite years of research in this area, it remains a time-consuming and challenging task due to schema mismatches, varying terminologies, and differences in data collection methodologies. This paper presents the case for agentic data harmonization as a means to both empower experts to harmonize their data and to streamline the process. We introduce Harmonia, a system that combines LLM-based reasoning, an interactive user interface, and a library of data harmonization primitives to automate the synthesis of data harmonization pipelines. We demonstrate Harmonia in a clinical data harmonization scenario, where it helps to interactively create reusable pipelines that map datasets to a standard format. Finally, we discuss challenges and open problems, and suggest research directions for advancing our vision.

cs.AI↗

arXiv2024

Sampling Methods for Inner Product Sketching

Recently, Bessa et al. (PODS 2023) showed that sketches based on coordinated weighted sampling theoretically and empirically outperform popular linear sketching methods like Johnson-Lindentrauss projection and CountSketch for the ubiquitous problem of inner product estimation. We further develop this finding by introducing and analyzing two alternative sampling-based methods. In contrast to the computationally expensive algorithm in Bessa et al., our methods run in linear time (to compute the sketch) and perform better in practice, significantly beating linear sketching on a variety of tasks. For example, they provide state-of-the-art results for estimating the correlation between columns in unjoined tables, a problem that we show how to reduce to inner product estimation in a black-box way. While based on known sampling techniques (threshold and priority sampling) we introduce significant new theoretical analysis to prove approximation guarantees for our methods.

cs.DB↗

arXiv2024

Simple Analysis of Priority Sampling

We prove a tight upper bound on the variance of the priority sampling method (aka sequential Poisson sampling). Our proof is significantly shorter and simpler than the original proof given by Mario Szegedy at STOC 2006, which resolved a conjecture by Duffield, Lund, and Thorup.

cs.DS↗

arXiv2024

Efficiently Estimating Mutual Information Between Attributes Across Tables

Relational data augmentation is a powerful technique for enhancing data analytics and improving machine learning models by incorporating columns from external datasets. However, it is challenging to efficiently discover relevant external tables to join with a given input table. Existing approaches rely on data discovery systems to identify joinable tables from external sources, typically based on overlap or containment. However, the sheer number of tables obtained from these systems results in irrelevant joins that need to be performed; this can be computationally expensive or even infeasible in practice. We address this limitation by proposing the use of efficient mutual information (MI) estimation for finding relevant joinable tables. We introduce a new sketching method that enables efficient evaluation of relationship discovery queries by estimating MI without materializing the joins and returning a smaller set of tables that are more likely to be relevant. We also demonstrate the effectiveness of our approach at approximating MI in extensive experiments using synthetic and real-world datasets.

cs.DB↗

arXiv2023

Weighted Minwise Hashing Beats Linear Sketching for Inner Product Estimation

We present a new approach for computing compact sketches that can be used to approximate the inner product between pairs of high-dimensional vectors. Based on the Weighted MinHash algorithm, our approach admits strong accuracy guarantees that improve on the guarantees of popular linear sketching approaches for inner product estimation, such as CountSketch and Johnson-Lindenstrauss projection. Specifically, while our method admits guarantees that exactly match linear sketching for dense vectors, it yields significantly lower error for sparse vectors with limited overlap between non-zero entries. Such vectors arise in many applications involving sparse data. They are also important in increasingly popular dataset search applications, where inner product sketches are used to estimate data covariance, conditional means, and other quantities involving columns in unjoined tables. We complement our theoretical results by showing that our approach empirically outperforms existing linear sketches and unweighted hashing-based sketches for sparse vectors.

cs.DB↗

arXiv2021

Auctus: A Dataset Search Engine for Data Augmentation

The large volumes of structured data currently available, from Web tables to open-data portals and enterprise data, open up new opportunities for progress in answering many important scientific, societal, and business questions. However, finding relevant data is difficult. While search engines have addressed this problem for Web documents, there are many new challenges involved in supporting the discovery of structured data. We demonstrate how the Auctus dataset search engine addresses some of these challenges. We describe the system architecture and how users can explore datasets through a rich set of queries. We also present case studies which show how Auctus supports data augmentation to improve machine learning models as well as to enrich analytics.

cs.IR↗

arXiv2021

Correlation Sketches for Approximate Join-Correlation Queries

The increasing availability of structured datasets, from Web tables and open-data portals to enterprise data, opens up opportunities~to enrich analytics and improve machine learning models through relational data augmentation. In this paper, we introduce a new class of data augmentation queries: join-correlation queries. Given a column $Q$ and a join column $K_Q$ from a query table $\mathcal{T}_Q$, retrieve tables $\mathcal{T}_X$ in a dataset collection such that $\mathcal{T}_X$ is joinable with $\mathcal{T}_Q$ on $K_Q$ and there is a column $C \in \mathcal{T}_X$ such that $Q$ is correlated with $C$. A naïve approach to evaluate these queries, which first finds joinable tables and then explicitly joins and computes correlations between $Q$ and all columns of the discovered tables, is prohibitively expensive. To efficiently support correlated column discovery, we 1) propose a sketching method that enables the construction of an index for a large number of tables and that provides accurate estimates for join-correlation queries, and 2) explore different scoring strategies that effectively rank the query results based on how well the columns are correlated with the query. We carry out a detailed experimental evaluation, using both synthetic and real data, which shows that our sketches attain high accuracy and the scoring strategies lead to high-quality rankings.

cs.DB↗

arXiv2019

Visus: An Interactive System for Automatic Machine Learning Model Building and Curation

While the demand for machine learning (ML) applications is booming, there is a scarcity of data scientists capable of building such models. Automatic machine learning (AutoML) approaches have been proposed that help with this problem by synthesizing end-to-end ML data processing pipelines. However, these follow a best-effort approach and a user in the loop is necessary to curate and refine the derived pipelines. Since domain experts often have little or no expertise in machine learning, easy-to-use interactive interfaces that guide them throughout the model building process are necessary. In this paper, we present Visus, a system designed to support the model building process and curation of ML data processing pipelines generated by AutoML systems. We describe the framework used to ground our design choices and a usage scenario enabled by Visus. Finally, we discuss the feedback received in user testing sessions with domain experts.

cs.LG↗

arXiv2019

A Topic-Agnostic Approach for Identifying Fake News Pages

Fake news and misinformation have been increasingly used to manipulate popular opinion and influence political processes. To better understand fake news, how they are propagated, and how to counter their effect, it is necessary to first identify them. Recently, approaches have been proposed to automatically classify articles as fake based on their content. An important challenge for these approaches comes from the dynamic nature of news: as new political events are covered, topics and discourse constantly change and thus, a classifier trained using content from articles published at a given time is likely to become ineffective in the future. To address this challenge, we propose a topic-agnostic (TAG) classification strategy that uses linguistic and web-markup features to identify fake news pages. We report experimental results using multiple data sets which show that our approach attains high accuracy in the identification of fake news, even as topics evolve over time.

cs.CL↗

arXiv2019

Bootstrapping Domain-Specific Content Discovery on the Web

The ability to continuously discover domain-specific content from the Web is critical for many applications. While focused crawling strategies have been shown to be effective for discovery, configuring a focused crawler is difficult and time-consuming. Given a domain of interest $D$, subject-matter experts (SMEs) must search for relevant websites and collect a set of representative Web pages to serve as training examples for creating a classifier that recognizes pages in $D$, as well as a set of pages to seed the crawl. In this paper, we propose DISCO, an approach designed to bootstrap domain-specific search. Given a small set of websites, DISCO aims to discover a large collection of relevant websites. DISCO uses a ranking-based framework that mimics the way users search for information on the Web: it iteratively discovers new pages, distills, and ranks them. It also applies multiple discovery strategies, including keyword-based and related queries issued to search engines, backward and forward crawling. By systematically combining these strategies, DISCO is able to attain high harvest rates and coverage for a variety of domains. We perform extensive experiments in four social-good domains, using data gathered by SMEs in the respective domains, and show that our approach is effective and outperforms state-of-the-art methods.

cs.IR↗