For data integration scenarios, make it possible to start a computation that uses as input more than just one dataset. This seems possible when defining an algorithm, which receives an array of DIDs.
This way, it would be possible to start a computation job that combines multiple datasets.
Another interesting application is data transformations. Allowing two datasets as input, it would be possible to have generic mapping algorithms that receive as input the data to be mapped and the mapping to be used. This also enables a market of data transformation mappings, which are shared and monetised as any other kind of dataset.
Solution
When configuring a computation, it is possible to define multiple DIDs that will be passed to the algorithm as inputs.
Motivation / Problem
For data integration scenarios, make it possible to start a computation that uses as input more than just one dataset. This seems possible when defining an algorithm, which receives an array of DIDs.
This way, it would be possible to start a computation job that combines multiple datasets.
Another interesting application is data transformations. Allowing two datasets as input, it would be possible to have generic mapping algorithms that receive as input the data to be mapped and the mapping to be used. This also enables a market of data transformation mappings, which are shared and monetised as any other kind of dataset.
Solution
When configuring a computation, it is possible to define multiple DIDs that will be passed to the algorithm as inputs.