At the moment, the final stage of the workflow uploads all extracted values to ORKG as literals. However, according to the original LLM Catalog comparison (https://orkg.org/comparisons/R1364660), extracted values fall into two categories: resources and literals. The current add_paper function should be extended to distinguish between these types and create ORKG resources for values that are intended to be resources, rather than storing them as literals. Only values corresponding to primitive data should remain literals.
At the moment, the final stage of the workflow uploads all extracted values to ORKG as literals. However, according to the original LLM Catalog comparison (https://orkg.org/comparisons/R1364660), extracted values fall into two categories: resources and literals. The current
add_paperfunction should be extended to distinguish between these types and create ORKG resources for values that are intended to be resources, rather than storing them as literals. Only values corresponding to primitive data should remain literals.