Elsevier is to use AI technology developed specifically for chemistry to extract and curate information such as images, drawings and reaction schemes from scientific publications and patents for its chemical database and search engine, Reaxys. It claims that substance information from images in patents and journals is captured far more quickly, accurately and at far greater scale than was previously possible.<\/p>\n
The technology developed by LG AI Research combines molecule detection, reaction-diagram parsing and optical chemical structure recognition (OCSR) in a single model. Called MolMole, it can extract machine-readable structures from images of chemical structures in scientific publications<\/a>.<\/p>\n \u2018Every hour a chemist spends deciphering figures or images to see what has already been made is an hour that could instead be spent on chemistry discovery,\u2019 says Mirit Eldor<\/a>, managing director, life sciences, at Elsevier. \u2018Our partnership with LG AI Research gives that time back, lifting more chemistry out of the image and into Reaxys \u2013 curated, searchable and ready to act on. A structure buried in a figure should be evidence rather than a dead end.\u2019<\/p>\n LG AI Research claims its technology outperforms its competitors at extracting chemistry from a full document page. To ensure accuracy, each extraction from a scientific publication is validated against existing Reaxys benchmarks before it goes live. The full pipeline has undergone rigorous testing across Elsevier\u2019s data and workflow tools before wider use, it adds.<\/p>\n \u2018We see a range of novel, partially improved AI models that compete for extracting chemistry with ever increasing quality,\u2019 says Lutz Weber<\/a>, co-founder of MolGenie a German software company. \u2018Given this great progress it is not astonishing that the data curation team of Elsevier has adopted one of those derivative methods such as MolMole. This should give curators at Reaxys a much higher productivity when compared with a fully manual curation.\u2019 However, there are still challenges, he adds. For example, the extraction of metal\u2013organic complexes or metal\u2013organic frameworks is not yet possible.<\/p>\n Weber notes LG\u2019s method \u2018unfortunately\u2019 can\u2019t be tested independently as the source code isn\u2019t available and there\u2019s no free webpage or service that allows for image-to-structure conversion. \u2018Thus, the reported superior testing results when compared with Decimer or MolScribe [similar, open access tools] cannot be validated. The paper also only appeared as an unreviewed arXiv publication.\u2019<\/p>\n