This repository contains the Xtract toolkit to automatically evaluate a collection of metadata attributes!
Input:
A folder containing one type of metadata extractor file. For example, it could be a folder full of tabular OR keyword metadata formatted as json.
Here we mine the following metrics:
- Completeness: out of the maximum possible fields in a metadata document, what percentage of them are full? The system automatically defines the 'maximum possible' as the total number of fields in the document with the highest number of fields.
- Entropy: TODO
- Yield: TODO
from quality_metric_tool import Summary
path = {path to folder of metadata}
summary = Summary()
all_scores = summary.get_quality_scores(path)