Data Smells in Public Datasets

Arumoy Shome

Luis Cruz

Arie van Deursen

This website consists of supplemental material for a research project.

[Publication][Pre-print]

1 Introduction

The adoption of Artificial Intelligence (AI) in high-stakes domains such as healthcare, wildlife preservation, autonomous driving and criminal justice system calls for a data-centric approach to AI. Data scientists spend the majority of their time studying and wrangling the data, yet tools to aid them with data analysis are lacking. This study identifies the recurrent data quality issues in public datasets. Analogous to code smells, we introduce a novel catalogue of data smells that can be used to indicate early signs of problems or technical debt in machine learning systems. To understand the prevalence of data quality issues in datasets, we analyse 25 public datasets and identify 14 data smells.

2 Data Analysis

Following is the list of 25 public datasets which were analysed to identify the data smells. The Manual analysis carried out for each dataset can be viewed by following the links below. Additionally, a meta analysis of the public datasets was conducted which is available on this page.

3 Data Smells Catalogue

The catalogue of data smells identified by this research project can be found on the catalogue page.