Text and Data Mining with AI

A new artificial intelligence (AI) system called LION LBD uses machine learning, natural language processing, and text mining to help researchers uncover links between data sets from millions of scientific studies. The literature-based discovery (LBD) technique was originally developed as a painstaking manual process in the 1980s, but computerization has made it a practical tool for finding patterns that humans would never have detected.

The first iteration of LION LBD is focused on cancer research due to the massive volume of research available across multiple scientific disciplines. The shear amount of data makes it impossible for scientists to keep up on all of the latest information, much less associate concepts from a variety of sources. LION LBD enables real-time search of tens of millions of publications and allows users to examine the data in its original context.

Currently, the system can only connect two keywords or concepts, but the developers have made the entire system open source and freely accessible to allow for collaborative development moving forward.

For information: Anna Korhonen, University of Cambridge, Language Technology Lab, 9 West Road, Cambridge CB3 9DP, United Kingdom; phone: +44-1223-767389; email: anna.korhonen@cl.cam.ac.uk; website: https://www.cam.ac.uk/