Neuronal e-mail classification at Fr. Meyer’s Sohn GmbH
Fr. Meyer’s Sohn (GmbH & Co.) KG relies on machine learning methods from interface projects GmbH for e-mail classification
Modern logistics thrives on optimised processes and closely networked information flows. The right data in the right place at the right time is one of the essential requirements within this industry.
For the Hamburg logistics company Fr. Meyer’s Sohn (FMS), e-mails are one of the central means of communication. The global player exchanges tens of thousands of messages daily between its own locations in more than 25 countries, service providers and customers. “We want to manage the manual process of e-mail communication with the means of artificial intelligence and were looking for a solution that autonomously recognizes content via neural networks and then classifies the mail thematically,” says Olaf Rathgeb, CTO at FMS.
The solution is based on an AI model trained with machine learning, which autonomously recognizes relevant entities within a message and classifies them accordingly in the context of the content. Unlike rule-based approaches, the system uses relationships in the text to capture the actual meaning. Based on a similarity search, a data-optimized model is created that robustly classifies the constant stream of new e-mails and then feeds them for further processing.
“We hope that this solution will enable faster communication and more efficient processing, as messages once classified can be delivered more quickly to the appropriate systems or processors. We are therefore looking forward to the completion of the pilot project, which we expect to confirm the viability of the solution,” Rathgeb continues. “In the future, neural networks will control the communication flow more intelligently and provide relevant information in a targeted manner at the right time, at the right place and in the right quantity”, adds Head of Sales Frank Kuckelkorn of interface projects GmbH.