An expanded partnership with Babelscape will bring new text analytics and smart tagging resources to Atex’s AI-powered publishing platform.

A new feature uses multilingual natural language processing to extract key concepts, entities and domains from text, and then recommend taxonomic labels to improve content discoverability.

A technology and innovation laboratory, Atex AI Platform has been established to develop use cases to exploit the potential of artificial intelligence in newsrooms. Among these, smart tagging allows publishing archives to be categorised automatically, saving time and resources.

Longtime partner Babelscape develops and commercialises innovative multilingual deep learning technologies. Co-founder and professor Roberto Navigli leads a team of developers that includes multiple PhD industrial engineers building multilingual semantic software to improve NL processing and understanding using techniques including machine and deep learning, symbolic AI and neural networks.

Navigli says the tools will not only add efficiency and cost savings to their editors work flows, but also increase the audiences for the content published on their CMS, by making it more discoverable through ‘smart tagging’.

Atex chief technology officer Michele Alessandrini said the partnership with Babelscape would allow them to “plug their experience into the workflows of our customers” by creating a feedback loop between the algorithms and the data.

“The result is to be able to offer a service not just able to interpret a language but to do so within the context of our publishers’ domains.”

Atex is a member of Toronto-based software and services provider Constellation Software. Babelscape’s partners and clients include Adobe, EUIPO, and non-profit research organisations and public institutions.