Consensus + Datasaur: A Revolutionary Science Answer Partnership
From Lennon and McCartney to Jobs and Wozniak, history-changing moments are often spurred by collaborators with an innate ability to see what’s next.
The collaboration between the science-based search product, Consensus, and NLP data-labeling platform Datasaur, may be among this lineage of history makers. Humbly speaking, of course.
With the help of Datasaur, our team at Consensus is building a revolutionary search engine that will give access to peer-reviewed scientific information to anyone who needs authoritative answers.
What Is Datasaur?
Datasaur was founded on the idea that building the right tools can further the adoption and democratization of artificial intelligence. Since its inception, Datasaur has become the leading NLP data-labeling platform working with the likes of Netflix and Zoom.
Datasaur’s platform allows its users to annotate, automate, integrate, isolate, and collaborate within their data. Thus, giving users a 10x reduction in project times and a 2x improvement in model performance. As the “most comprehensive labeling tool in the market,” Datasaur strives to create a machine learning platform accessible to companies both large and small.
How Do Consensus and Datasaur Work Together?
In order for our state-of-the-art machine learning system to effectively analyze and extract information from scientific literature, we needed to construct quality datasets to train our models. To build those datasets, we needed an annotation product that:
- Labels scientific papers at different levels (entity, sentence, multiple sentences, etc.)
- Allows us to create custom label sets
- Enables multiple annotators in the same file
- Supports inexperienced annotators (in our case, Ph.D.’s and Ph.D. students)
- Provides easy to use workflow tools
Datasaur met our complex annotation requirements and more. After assuring we were set up correctly within their UI, Datasaur trained us on best practices to maximize functionality. With this collaboration, we are expediting the indexation and data quality of our search engine.
This early success enabled a second project and workflow with Datasaur that utilizes a different technique of labeling and review.
Sign up for Our Beta
Read more about the exciting work to come from our partnership with Datasaur. And if you’re interested in gaining access to peer-reviewed scientific research before anyone else, sign up for our free beta.