Semantic expansion of bibliographic records

A visualization paradigm based on geolocated data. See online demo here!

What's this?

This is the project I made for my CS thesis. In short:

• Pulls and parses bibliographic data from an OAI-PMH based endpoint (here)
• Uses entity extraction tasks (spaCy) based on ML to generate semantic links
• Adds Wikipedia annotations and Wikidata SPARQL queries to enrich data
• Displays results in a web application based on a dynamic interactive map (Leaflet)

Prerequisites

• Python 3 installed (tested on v3.6)

Requirements

• lxml, beautifulsoup4, wikipedia, sickle, flask, spacy
• Spacy's italian statistical model (python -m spacy download it_core_web_sm), see more here

How to run

Run FLASK_APP=app.py flask run to launch the Web Server and visit localhost:5000