Ivan Vankov

Cognitive Scientist & Software Engineer at JSL

Ivan Vankov has been doing research in the cognitive sciences for more than 10 years. His main research interest is how knowledge is represented in biological and artificial neural networks. He currently works as a machine learning engineer for John Snow Labs, where is responsible for implementing and integrating state-of-the-art deep learning models for natural language processing.

WATCH LIVE: February 17 @ 1:50PM – 2:10PM ET

State-of-the-art relation extraction from unstructured clinical text

One of the most challenging NLP tasks is identifying relations between textual entities. Traditional approaches, based on feature engineering, usually generalize poorly to novel data sets. In this talk I will present a novel approach which utilizes a state-of-the-art biomedical language representation deep learning model. The model is fully integrated into the SparkNLP library and makes use of other components such as named entity recognition and dependency parsing. I will demonstrate how SparkNLP is used to build a complete NLP pipeline for identifying and classifying relations in raw clinical texts.