You can compare the extracted results from the
same input file by using the Annotation Difference Viewer. You can
also compare the results of one extractor run to a labeled data
collection to evaluate the quality of an extractor.
Comparing
the results of one extractor run to the results of another run
Comparing the results of one extractor run to the results of another
run can be useful when you are testing an extractor. Each time a
launch configuration is run, a new subdirectory named result-system-timestamp
is saved in the result directory of
the project.
Creating
a labeled data collection A labeled
data collection is the standard way to evaluate the accuracy and
completeness of an extractor.
Configuring
the labeled data collection The labeled
document editor is a text editor in which you can manually label
documents. Before you can label a document, you must configure the
collection of which the document is a part.