Computing spectral sequences (Q2457386)

As a computational tool spectral sequences suffer from the fact that although one can compute the chain groups of the \(E^{r}\) term from the \(E^{r-1}\) terms and \(d_{r-1}\), the determination of the differential \(d_{r}\) requires additional insight. This paper uses examples to explain how the notion of an ``object with effective homology'' allows one to determine algorithmically not only the \(E^{r}\) terms of a spectral sequence but also the differentials. The paper concentrates on descriptions of illustrative examples that use the Kenzo system -- a homological algebra computation system written in LISP -- to compute examples of the Serre and Eilenberg-Moore spectral sequences.