Calculating effect of entity recognition on a relation extraction system

How can we calculate/formulate the effectiveness of named entity linking (based on P/R/F1 or any other evaluation metrics) on a relation extraction system which accepts the output of ER as its input? Suppose we have a serial architecture in which entity recognition (NER or EL) performs before relation extraction, and RE uses the entities which are extracted by ER module. Note that, relation extraction module just detects the relationship between each pair of entities in the sentence.

I found this article:

Relation Extraction and the Influence of Automatic Named-Entity Recognition, by CLAUDIO GIULIANO, ALBERTO LAVELLI, and LORENZA ROMANO

But I would like to ensure we cannot formulate the dependency between two modules in my case of study.

Topic information-extraction named-entity-recognition evaluation classification

Category Data Science

About

Geeks Mental is a community that publishes articles and tutorials about Web, Android, Data Science, new techniques and Linux security.