Leveraging Large Language Models to Identify Similar Data Points Across Diverse Datasets
Our study introduces a novel two-step approach that leverages Large Language Models (LLMs) to efficiently identify similar data points across diverse, non-free text domains such as tabular and image data.
Automated Insight Discovery and Exploration in Large-Language-Model-Powered Data Analysis
A multi-agent framework that automatically extracts, associates, and organizes insights from conversational contexts to facilitate efficient insight discovery and exploration during LLM-powered data analysis.
ℓ1-Norm Regularized ℓ1-Norm Best-Fit Lines: Optimization Framework for Sparse Robust Subspace Estimation
Proposing an optimization framework for estimating a sparse robust one-dimensional subspace using ℓ1-norm regularization.
ClusterNet: A Perception-Based Clustering Model for Scattered Data Analysis
ClusterNet outperforms existing clustering techniques in aligning with human judgments and demonstrates the ability to generalize to unseen data.
SheetAgent: A Generalist Agent for Spreadsheet Reasoning and Manipulation via Large Language Models
SheetAgent demonstrates superior spreadsheet manipulation and reasoning capabilities using LLMs.
Pareto-Optimal Estimation and Policy Learning for Short-term and Long-term Treatment Effects
Developing Pareto-optimal estimation and policy learning to identify the most effective treatment that maximizes total reward from short-term and long-term effects.
G-Mapper: Optimizing Cover Parameter in Mapper Construction
G-Mapper optimizes cover parameter using G-means clustering for Mapper construction.
ForTune: Running Offline Scenarios to Estimate Impact on Business Metrics
Product leaders face challenges in making informed decisions, but ForTune offers a novel approach to investigate hypotheses and predict business metrics accurately.
Umwelt: Accessible Structured Editing of Multimodal Data Representations
Umwelt provides an authoring environment for interactive multimodal data representations, de-centering the visual modality and emphasizing equal representation of visualization, sonification, and textual description.
BIBench: Evaluating Large Language Models for Data Analysis in Business Intelligence
Large Language Models (LLMs) are assessed for their data analysis capabilities in Business Intelligence through BIBench, focusing on BI foundational knowledge, knowledge application, and technical skills.
