Neo4j-based Graph Database of WordNet Type, built on distributional data from Russian Distributional Thesaurus
In this project for the "Databases 2" course we, as a team of students:
- parced the information about frequency of hyponimic relations between a set of words
- parced the connections between the meanings of the same set of words
- invented the architecture for our database
- optimised parced data for the DB creation with Neo4j methods and instruments
- programmed a custom web-interface for interaction with the DB
The performance of the DB was tested on:
- CRUD actions
- Sorting, grouping and clusterisation
- Path finding for distant nodes