####0. Preparation

• Download Wikidata/Wikipedia dumps.
• Import Wikipedia *.sql to MySQL database.
• Prepare Stanford Tokenizer and Corenlp.
• Prepare WikiExtractor.

####1. Extract all items, properties, triples from Wikidata. python wd.extract_all.py target_dir

####2. Tokenize the entities' description, aka. python wd.tokenize.py

####3. Extract linked sentences from Wikipedia xml dump parsed by WikiExtractor.py. python wp.extract_all.py

####4. Combine Two data. python wp.combine_wd.py

####5. Create wikiP2D dataset. python create_dataset.py