D4.3 – Evaluation Set-Up

This document provides an overview of the planned evaluation of NLP pipelines for various languages. Two forms of evaluation were planned, an intrinsic and an extrinsic one, the former intended to provide direct insight into the performance of our named entity recognition and time expression identification methods on biographical data in comparison to human-annotated data. The extrinsic evaluation was intended to study the possibility of using the output of our tools in specific research questions.

This deliverable is currently classified as confidential and is not available for download.