EPL-Knowledge-Graph-and-Game-Prediction

DSCI558 Building Knowledge Graph Project

1. Crawler

WorldFootball.net (https://www.worldfootball.net)

cd KG_generation/worldfootball_crawler
game: `scrapy crawl game -o game_worldfootball.jl'
player_name: 'scrapy crawl player -o player_worldfootball.jl'
player_info: 'scrapy crawl player_info -o player_info_worldfootball.jl'
team_info: 'scrapy crawl team_info -o team_info_worldfootball.jl'

Whoscored.com

cd ../whoscored_crawler
run whoscored_Crawler.ipynb

Wikipedia

cd ../wikipedia_crawler
scrapy crawl team_info team_info_wikipedia.jl

2. Entity Resolution

cd ../Entity Resolution
run 1. Player Entity Resolution.ipynb
run 2. Player PSL.ipynb
run 3. Team Entity Resolution.ipynb
run 4. Game Entity Resolution.ipynb

3. KG Construction

cd ../KG Construction
run 5. KG Construction.ipynb

4. Neo4j

1. Install Neo4j Desktop -- w/ Neosemantics Plugin.
2. Upload RDF file (data.ttl) into Neo4J desktop.
3. Run Neo4J desktop.

DATABASE_USERNAME="neo4j"
DATABASE_PASSWORD="dsci558!"
DATABASE_URL="bolt://localhost:7687"

5. Flask

cd backend
conda create -n neo4j python=3.7
conda activate neo4j
pip install -r requirements.txt
export FLASK_APP=app.py
flask run

FLASK_URL="http://localhost:5000"

6. Frontend

npm install
npm run serve

5. Contributor

• 1. Jae Young Kim : [jkim2458@usc.edu]
• 2. Dong Ho Lee : [dongho.lee@usc.edu]