Tripl\`etoile: Extraction of Knowledge from Microblogging Text

TL;DR

This paper presents Tripl extbackslash etoile, a novel pipeline for extracting open-domain knowledge graphs from microblogging posts, achieving high precision and outperforming existing methods on Twitter data.

Contribution

The paper introduces an enhanced extraction pipeline that uses dependency parsing and unsupervised relation classification tailored for social media text.

Findings

Achieves over 95% precision in extracting triples.

Outperforms similar pipelines by approximately 5% in precision.

Generates a higher number of triples compared to existing methods.

Abstract

Numerous methods and pipelines have recently emerged for the automatic extraction of knowledge graphs from documents such as scientific publications and patents. However, adapting these methods to incorporate alternative text sources like micro-blogging posts and news has proven challenging as they struggle to model open-domain entities and relations, typically found in these sources. In this paper, we propose an enhanced information extraction pipeline tailored to the extraction of a knowledge graph comprising open-domain entities from micro-blogging posts on social media platforms. Our pipeline leverages dependency parsing and classifies entity relations in an unsupervised manner through hierarchical clustering over word embeddings. We provide a use case on extracting semantic triples from a corpus of 100 thousand tweets about digital transformation and publicly release the generated knowledge graph. On the same dataset, we conduct two experimental evaluations, showing that the system produces triples with precision over 95% and outperforms similar pipelines of around 5% in terms of precision, while generating a comparatively higher number of triples.