How FactSet automated entity matching against official registry data

FactSet has built a model that automatically matches the companies in its database to official registry records on OpenCorporates, as part of a larger strategy to ensure their legal entity data coverage is updated and accurate. It uses software to achieve what a researcher used to do by hand: search for a company, work through the candidates that come back, and identify the one official record that is genuinely the right match. Across several thousand test companies, it reaches match rates of up to around 70 percent, with almost no false positives.