liveBook · Manning

Data mining

As explained in Introduction to Generative AI, Second Edition

Data mining is the computational analysis of large collections of data to discover patterns, information, or relationships.

Data mining uses computers to examine large collections of data for patterns or relationships. When the data includes creative work, a practical question follows: who may use that work, and under what conditions? German copyright law permits data mining when data is lawfully accessible and subsequently deleted, but generative models have brought such rules under greater scrutiny. section 4.11

Images in a large dataset

LAION-5B contains links to billions of publicly available images and is used by developers of generative image models. A photographer asked for his stock images to be removed. LAION said it could not readily identify them in its link-only database; Stability AI later announced it would honor artists’ opt-out requests. The dispute shows how difficult it can be to identify and address individual contributions at this scale. section 4.11

How Introduction to Generative AI, Second Edition covers it

How deep it goes
States a German legal condition and describes a removal dispute, without explaining mining algorithms or resolving the dispute’s legality.
Examples
Uses LAION-5B, a photographer’s removal request, and Stability AI’s opt-out announcement.
What you can do after
Evaluate/understand why mining creative work raises questions about access, removal, and artists’ control.
How it connects
Connects data mining to copyright law, image datasets, generative models, and artist opt-outs.

Also explained in

Compare how 4 books explain data mining →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Introduction to Generative AI, Second Edition. The section links point to the parts of the book it draws on.