Team showcase how good Watson is at learning

IBM supercomputer WatsonArtificial intelligence has undoubtedly come a long way in the last few years, but there is still much to be done to make it intuitive to use.  IBM’s Watson has been one of the most well known exponents during this time, but despite it’s initial success, there are issues to overcome with it.

A team led by Georgia Tech are attempting to do just that.  They’re looking to train Watson to get better at returning answers to specific queries.

Training Watson

The training consisted of around 1,200 question-answer pairs that allowed the team to chat to Watson and subsequently gain inspiration for design challenges in areas such as engineering and computing.

The team worked alongside Watson to devise solutions that could take their inspiration from the natural world.  They did this by feeding Watson a few hundred biology articles from the interactive biology repository Biologue.  The system was then quizzed on what it had learned.

Watson was quickly able to pluck answers to a wide range of questions to guide the team through the innovation process.  They were therefore able to gain access to a much wider pool of knowledge than would otherwise have been the case.

The team believe that the ability for Watson to retrieve natural language information allows relative novices to quickly train Watson on fairly complex topics.

“Researchers are provided a quickly digestible visual map of the concepts relevant to the query and the degree to which they are relevant,” they say. “We were able to add more semantic and contextual meaning to Watson to give some notion of a conversation with the AI.”

The next step is for the team to explore how effective Watson could be in other fields, such as healthcare or education.


Leave a Reply

Your email address will not be published. Required fields are marked *