Monday Morning Data Chat #99 - Dealing with Dirty Data and Data Hubris w/ Jessica Talisman
7 snips
Oct 24, 2022 
Taxonomist Jessica Talisman explores why clean, useful data takes more than trendy algorithms. From tangled e-commerce categories and emojis to speed bumps with three different names, the conversation digs into how language, culture, and human expertise shape data systems. They also touch on machine learning, knowledge graphs, UX, data quality, and the risks of charging ahead without understanding the data.
Chapters
Transcript
Episode notes
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Introduction
00:00 • 2min
Is There a Poly Hierarchy in E-Commerce?
02:20 • 3min
How Do You Manage Multiple Relationships in Taxonomies?
05:13 • 5min
Data Catalog
10:16 • 4min
Is There a Difference Between Language and Music Classifications?
13:48 • 2min
Is There a Difference Between Ontology and Taxonomy?
16:08 • 3min
Master Data Management - Controlled Vocabulary and Crosswalks
18:44 • 3min
Data Engineers - What Type of Data Engineer Should We Hire?
21:30 • 2min
Is ML Clustering a Good Addition to Human Driven Taxonomy?
23:11 • 4min
NLP and Taxonomy
26:44 • 3min
Is Stable Diffusion a Good Example of an Industry Standard Taxonomy?
29:55 • 4min
Do You Know Some Latin?
33:54 • 2min
Do Emojis Have a Taxonomy?
36:07 • 3min
Symbols for Efficiency and Speed
39:23 • 2min
UX and UX Folk Design Products Work Closely Together
41:46 • 5min
Taxonomy and the Data Quality Dashboard
46:17 • 4min
Data Engineering and Data Science
50:45 • 2min
The Value of Copy Cataloging in Data Management
53:14 • 5min
How to Standardize Ethics and Bias?
57:56 • 2min
What's Up With You?
59:37 • 2min