Publishing chemistry data in a way that allows computers to actually use it
What is it about?
Chemistry research articles might as well be published in the 19th century: the way molecules and data are prepared for manuscripts does not allow computers to understand or use it. Even when computers are used to prepare documents, it is designed only to be viewed by people. Chemists need to start using modern data capture methods in order that machine learning methods can use their research.
Why is it important?
Open lab notebooks are becoming increasingly popular, which means that the amount of data will continue to expand. The risk is that if traditional publication methods continue to be used, this will mean that there is an exponentially increasing quantity of bad data, rather than a groundbreaking opportunity to use computers to learn from previous experiments.
The following have contributed to this page: Dr Antony John Williams, Alex Michael Clark, and Dr Sean Ekins
In partnership with: