Unsupervised context retrieval for long-tail entities

This talk presented our work in the context of monitoring entities, like persons and companies, in news streams. We study the problem of retrieving, in an unsupervised fashion, textual contexts that can support the acquisition of labeled data to train machine learning models for entity linking.

We developed this project during my internship at Signal AI in London.
This work was accepted as an article, and presented, at the 2019 ACM SIGIR International Conference on Theory of Information Retrieval (ICTIR ‘19), which took place in Santa Clara, CA, USA.