Freebase is a large collaborative knowledge base consisting of metadata composed mainly by its community members. It is an online collection of structured data harvested from many sources, including individual 'wiki' contributions. Freebase aims to create a global resource which allows people (and machines) to access common information more effectively.
Freebase data is available for free/libre for commercial and non-commercial use under a Creative Commons Attribution License, and an open API, RDF endpoint, and database dump are provided for programmers. Freebase contains 1.2 billion of triples so having this information in a graph will be very useful in order to be able to create new graph-based algorithms for disambiguation.
The goal of this task is to develop a tool that will be able to parse a Freebase dump and import it in a graph database. For a first version, the selected database will be Neo4j, managed by Tinkerpop Blueprints API. In order to build the graph, we are going to store as vertexes all the Freebase entities. Between two vertex, it will be an edge if there is a direct relationship between both entities in Freebase or a mediated relationship. Mediated relationship relates two entities that are Topics (concepts) in Freebase through an entity that is not a Topic (categories). This categorized edges can be very useful for the later disambiguation.