Seed.aml should be generated or taken from already data

[omarrana]

For the generation of heterogeneities, we need seed.aml as an input which has the possibility to be split into documents d1,d2...dn containing heterogeneities.

Should we also generate seed.aml or our input would have a manually created seed.aml by human expert containing possibility for heterogeneities.

We already have 200-400 testbeds containing seed.aml.

[omarrana]

close it please

[igrangel]

The generation of the heterogeneities was one of the criticized points. This is not completely clear right now. Here is important to describe, how the heterogeneities are generated and under which assumptions they are generated.