MetaInsight: Automatic Discovery of Structured Knowledge for Exploratory Data Analysis

Ma, Pingchuan; Ding, Rui; Han, Shi; Zhang, Dongmei

doi:10.1145/3448016.3457267

Cited by 21 publications

(13 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Insight exploration methods have been demonstrated to be useful in deeper data analysis. 54–56 Such findings indicate that we may see similar advances with counterfactuals, that is, improved analyses and causal inferences.…”

Section: Methodsmentioning

confidence: 83%

An empirical study of counterfactual visualization to support visual causal inference

Wang,

Borland,

Gotz

2024

Information Visualization

View full text Add to dashboard Cite

Counterfactuals – expressing what might have been true under different circumstances – have been widely applied in statistics and machine learning to help understand causal relationships. More recently, counterfactuals have begun to emerge as a technique being applied within visualization research. However, it remains unclear to what extent counterfactuals can aid with visual data communication. In this paper, we primarily focus on assessing the quality of users’ understanding of data when provided with counterfactual visualizations. We propose a preliminary model of causality comprehension by connecting theories from causal inference and visual data communication. Leveraging this model, we conducted an empirical study to explore how counterfactuals can improve users’ understanding of data in static visualizations. Our results indicate that visualizing counterfactuals had a positive impact on participants’ interpretations of causal relations within datasets. These results motivate a discussion of how to more effectively incorporate counterfactuals into data visualizations.

show abstract

Section: Methodsmentioning

confidence: 83%

An empirical study of counterfactual visualization to support visual causal inference

Wang,

Borland,

Gotz

2024

Information Visualization

View full text Add to dashboard Cite

show abstract

“…Then she might forward this insight and investigate the root-cause of delays in these airports further. Please note: there is a lack of explicit intents (Milo et al 2020;Ma et al 2021a) Table 1: Sampling strategies and corresponding parameters that together creates our action space. We use the Short Name to refer these.…”

Section: Motivating Examplementioning

confidence: 99%

“…a query) against a tabular data, receives some answers, subsection of the tabular data or some visualization and then decides which queries to issue next in order to understand the hidden characteristics of the data and associated insights better (Ma et al 2021a;Bar El et al 2020;Milo et al 2018a). Characteristics and insights obtained using EDA are crucial for subsequent decision making in various domains (Ma et al 2021a). The success of an EDA session and the quality of insights obtained from it large depend on two things: (1) how interactive the system is (i.e.…”

Section: Introductionmentioning

confidence: 99%

Reinforced Approximate Exploratory Data Analysis

Garg

Mitra

Tao

et al. 2023

AAAI

View full text Add to dashboard Cite

Exploratory data analytics (EDA) is a sequential decision making process where analysts choose subsequent queries that might lead to some interesting insights based on the previous queries and corresponding results. Data processing systems often execute the queries on samples to produce results with low latency. Different downsampling strategy preserves different statistics of the data and have different magnitude of latency reductions. The optimum choice of sampling strategy often depends on the particular context of the analysis flow and the hidden intent of the analyst. In this paper, we are the first to consider the impact of sampling in interactive data exploration settings as they introduce approximation errors. We propose a Deep Reinforcement Learning (DRL) based framework which can optimize the sample selection in order to keep the analysis and insight generation flow intact. Evaluations with real datasets show that our technique can preserve the original insight generation flow while improving the interaction latency, compared to baseline methods.

show abstract

“…A variable is either categorical or numerical. Following the conventions in data analysis [13,25], we denote a categorical variable as dimension and a numerical variable as measure. For relational data, we anticipate to take a materialized provenance table [23] as the input to our system.…”

Section: Preliminary 21 Data Model and Querymentioning

confidence: 99%

“…Exploratory data analysis (EDA) is a key step for acquiring insight from data and facilitating analysis towards decision making [25,33]. With the advent of the digital age, the information explosion phenomenon [10] makes it difficult for users to justify and rely on knowledge and conclusions from EDA.…”

Section: Introductionmentioning

confidence: 99%

XInsight: eXplainable Data Analysis Through The Lens of Causality

Ma¹,

Ding²,

Wang³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

In light of the growing popularity of Exploratory Data Analysis (EDA), understanding the underlying causes of the knowledge acquired by EDA is crucial, but remains under-researched. This study promotes for the first time a transparent and explicable perspective on data analysis, called eXplainable Data Analysis (XDA). XDA provides data analysis with qualitative and quantitative explanations of causal and non-causal semantics. This way, XDA will significantly improve human understanding and confidence in the outcomes of data analysis, facilitating accurate data interpretation and decision making in the real world. For this purpose, we present XInsight, a general framework for XDA. XInsight is a three-module, endto-end pipeline designed to extract causal graphs, translate causal primitives into XDA semantics, and quantify the quantitative contribution of each explanation to a data fact. XInsight uses a set of design concepts and optimizations to address the inherent difficulties associated with integrating causality into XDA. Experiments on synthetic and real-world datasets as well as human evaluations demonstrate the highly promising capabilities of XInsight.

show abstract

MetaInsight: Automatic Discovery of Structured Knowledge for Exploratory Data Analysis

Cited by 21 publications

References 25 publications

An empirical study of counterfactual visualization to support visual causal inference

An empirical study of counterfactual visualization to support visual causal inference

Reinforced Approximate Exploratory Data Analysis

XInsight: eXplainable Data Analysis Through The Lens of Causality

Contact Info

Product

Resources

About