Current State-of-the-Art Accuracy for Key Medical Natural Language Processing Benchmarks

Being the most widely used library in the healthcare industry, Spark NLP for Healthcare comes with 700+ pretrained clinical models that are all developed & trained with latest SOTA algorithms to solve real world problems in healthcare domain at scale.

In order to provide accurate and reliable models and tools all the time while covering the edge cases in real word scenarios and to improve the generalisation power of the DL models, the datasets and models are monitored, augmented and updated on a regular basis so that they can be used out of the box with no further efforts.

In this talk, Veysel will share the latest accuracy benchmarks from the healthcare-specific models of Spark NLP library (De-Identification, Named Entity Recognition and Entity Resolution Models) with respect to the academic benchmarks published by researchers and the commercial solutions provided by major cloud providers (AWS Medical Comprehend, GCP Healthcare API and Azure Text Analytics for Health).