Artificial Intelligence Delivers Real Results
La Plante, Louis, PM Network
Researching a chemical compound for a major pharmaceutical project is normally a months-long, laborious process that requires legions of scientists, patent engineers and attorneys. Global computer giant IBM is out to make that onerous task not only faster, but also cheaper.
Six years in the making, the company's strategic internet protocol insight platform (SIIP) could represent the next generation of artificial intelligencedriven computing. SIIP's algorithms extract information from texts and images and normalize the data for project researchers, searching and compiling thousands of compounds in mere seconds.
Before launching SIIP last December, IBM collected data from 4.7 million patents and 11 million abstracts from biomedical journals published between 1976 and 2000. That raw data has long been available to researchers worldwide, but the complexity of chemical compounds and the lack of a common language within the industry previously made large-scale compilation prohibitive. One chemical compound, for example, could have more than 100 different names in thousands of medical journals. …