I have read a lot of papers about constructing a minimal model for a dga $(A,d)$, but still struggle to summarise the minimal data I need to construct the minimal model. Can someone help!
Also, what information does the minimal model captures that the cohomology doesn't?