Archive

Archive for September, 2012

Paper: Comparison of named entity recognition tools for raw OCR text

This short paper analyses an experiment comparing the efficacy of several Named Entity Recognition (NER) tools at extracting entities directly from the output of an optical character recognition (OCR) workflow. The authors present how they first created a set of test data, consisting of raw and corrected OCR output manually annotated with people, locations, and organizations. They then ran each of the NER tools against both raw and corrected OCR output, comparing the precision, recall, and F1 score against the manually annotated data.

Read the paper in the online proceedings of the KONVENS 2012

Advertisements
Categories: EHRI