Integroly: Automatic Knowledge Graph Population from Social Big Data in the Political Marketing Domain

Abstract

Social media sites have become platforms for conversation and channels to share experiences and opinions, promoting public discourse. In particular, their use has increased in political topics, such as citizen participation, proselytism, or political discussions. Political marketing involves collecting, monitoring, processing, and analyzing large amounts of voters’ data. However, the extraction, integration, processing, and storage of these torrents of relevant data in the political domain is a very challenging endeavor. In the recent years, the semantic technologies as ontologies and knowledge graphs (KGs) have proven effective in supporting knowledge extraction and management, providing solutions in heterogeneous data sources integration and the complexity of finding meaningful relationships. This work focuses on providing an automated solution for the population of a political marketing-related KG from Spanish texts through Natural Language Processing (NLP) techniques. The aim of the proposed framework is to gather significant data from semi-structured and unstructured digital media sources to feed a KG previously defined sustained by an ontological model in the political marketing domain. Twitter and political news sites were used to test the usefulness of the automatic KG population approach. The resulting KG was evaluated through 18 quality requirements, which ensure the optimal integration of political knowledge.