Triplètoile: Extraction of Knowledge from Microblogging Text
This work addresses the problem of adapting knowledge extraction to open-domain microblogging sources for researchers and practitioners in natural language processing, representing an incremental improvement over existing methods.
The paper tackles the challenge of extracting knowledge graphs from microblogging text, such as tweets, by proposing an enhanced pipeline that uses dependency parsing and unsupervised hierarchical clustering, achieving over 95% precision and outperforming similar methods by about 5% on a dataset of 100,000 tweets.
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.