Tech giant IBM is making the technology that underpins its AI debating system, Project Debater, commercially available to enterprises to help them understand the language of business.
Project Debater’s natural language processing (NLP) algorithms will be integrated with IBM’s AI platform, Watson.
It will see Project Debater’s sentiment analysis feature used to identify and analyse colloquialisms such as “hot under the collar”, which have historically been difficult for AI systems to spot.
“Language is a tool for expressing thought and opinion, as much as it is a tool for information,” said Rob Thomas, general manager, IBM Data and AI.
“This is why we’re harvesting technology from Project Debater and integrating it into Watson – to enable businesses to capture, analyse, and understand more from human language and start to transform how they utilise intellectual capital codified in data.”
Project Debater rolled out to Watson in three phases
IBM will make Project Debater’s NLP abilities available to businesses through an application programme interface (API).
Big Blue will roll out different capabilities in stages. Project Debater’s Advanced Sentiment Analysis technology, which aims to identify idioms, will be added to Watson this month.
Summarisation technology, which trawls through millions of text and audio to provide a summary of what is being said about a topic, will be added to Watson Natural Language Understanding later this year.
And Project Debater’s Advanced Topic Clustering, which gathers incoming data into meaningful clusters – such as recognising language used by specific industries like insurance and healthcare – will be integrated with Watson Discovery later on in the year.
IBM launched Project Debater in 2012. It is designed to construct speech on any topic based on information it is fed. It can then identify arguments made by a human debater and provide rebuttals.
In February 2019 Project Debater impressed during a debating contest with one of the world’s top competitive debaters in the world.
While it did not win the debate outright, Verdict reported at the time that it “gave a very strong and coherent performance, showing its ability to formulate an argument, bolster that argument with information and present it in a fluid and articulate way”.