San Diego, CA – Parity Computing today announced the beta launch of Parity DataSense, a scientific dataset discovery system with unique citation linking features. Parity DataSense users can query multiple dataset repositories simultaneously. Result ranking is influenced by dataset citations from the full-text literature, to help users identify high-quality datasets that have been repeatedly reused. The beta version of Parity DataSense is freely available to the public at:
http://www.paritydatasense.com

There is growing recognition that publicly-available high-quality datasets are crucial for reproducibility and for the productivity of data-intensive science. A rapidly increasing number of datasets are available from dataset repositories. But search and analysis across disparate repositories remain challenging, and Parity DataSense addresses this challenge. “We are leveraging literature citations to show users how other researchers have successfully reused specific datasets,” said Chris Rosin, Parity’s Chief Scientist.

While research articles generally identify the datasets they use, this is often done in an informal way embedded in the midst of article full-text rather than in the bibliography. Parity is adapting its industry-leading entity resolution technology to extract and link these dataset citations with high accuracy. Michaeleen Trimarchi, Parity’s Chief Content Analyst and former Library Manager at Scripps Research, said “Parity’s technological approach is an important complement to other community initiatives that seek to strengthen the linkage between research datasets and literature.”

The public beta of Parity DataSense is available today, and its features, repositories, and literature coverage will grow over time. Parity invites feedback from the community to help shape the development of Parity DataSense.

About Parity Computing: Parity provides powerful data mining, analytics, and decision-support systems to publishers and institutions in science, technology, and medicine (STM).