Contains scripts for PropertySuggester to preprocess the wikidata dump

• use dumpconverter.py to convert a wikidata dump to csv
• use analyzer.py to create a csv file with the suggestion data that can be loaded into a sql table
• the PropertySuggester extension provides a maintenance script (maintenance/UpdateTable.php) that allows to load the csv into the database

python dumpconverter.py wikidatawiki-20140226-pages-articles.xml.bz2 dump.csv
python analyzer.py dump.csv wbs_propertypairs.csv
php extensions/PropertySuggester/maintenance/UpdateTable.php wbs_propertypairs.csv

sudo apt-get install build-essential python-pip python-dev
pip install -r requirements.txt

nosetests

• Consider classifying Properties
• use Json dumps for analysis

• Generate associationrules for qualifier and references
• Improve ranking to avoid suggestions of human properties
• remove very unlikely rules (<1%)

• Converts a wikidata dump to a csv file with associationrules between properties