John Snow Labs, an award-winning global data operations and AI company, today announces that its Natural Language Understanding Software Spark NLP has gained recognition in a three-industry event: Spark+AI Summit, Predictive Analytics World Healthcare, and Strata Data.
Since its first public release nine months ago, the John Snow Labs NLP Library for Apache Spark has experienced widespread adoption as it set a new bar for production-grade natural language understanding. In terms of performance, detailed benchmarks published by O’Reilly Media show the library to be 38x to 80x times faster than spaCy (the top performing library to date) on a single machine. Spark NLP is also the only open source library which can natively scale on a distributed cluster, often delivering near-linear scalability.
In terms of accuracy, Spark NLP is the first and only library to productize several of the most recent and best performing deep learning algorithms for NLP. Among others, the library provides scalable, production-grade code for named entity recognition, assertion status detection, and entity resolution, based on academic papers that were published less than a year ago. The library’s API enables users to either leverage pre-trained models or train their own models for tackling domain-specific texts.
As a result, three of the most prominent and selective technical conferences in the AI space have recently chosen to highlight Spark NLP as the leading technology choice for data scientists:
* Spark+AI Summit, June 2018 in San Francisco, featured a “State of the Art Natural Language Processing at Scale” technical deep dive, focused exclusively on Spark NLP
* Predictive Analytics World Healthcare, June 2018 in Las Vegas, featured a session on “Natural Language Understanding in Healthcare: State of the Art NLP, Machine Learning and Deep Learning with Open Source Software“, focused exclusively on Spark NLP
* Strata Data, September 2018 in New York, will this year convert its half-day tutorial on “Natural Language Understanding at Scale” to cover only Spark NLP, and some of its advanced use cases
“We are super excited to see how quickly the industry’s most selective thought leaders have picked up Spark NLP and increasingly recommend it as the default choice to data scientists worldwide. We are committed as ever to keep pushing the state of the art and provide the community with the best performing and most accurate NLP software ever built”, said Saif Addin Ellafi, lead NLP engineer at John Snow Labs.
John Snow Labs maintains a full-time development team to keep improving the open source library, which delivered 10 new releases during the first six months of 2018. The library is monetized by licensing pre-trained models and data sets for the healthcare vertical. The models and data sets are continuously updated and optimized for accuracy on top of the high-performing Spark NLP core. The global market for natural language pro.