Last year, Thad Lurie, senior vice president of digital and technology at the American Geophysical Union (AGU), began looking into ways to use AI to deliver personalized information, education, and connection recommendations to members. "We have literally hundreds of thousands of peer-reviewed articles and scientific abstracts," he said. "People were saying, 'We know the information is there, we just can't find it.'"
Through a technology partner, AGU developed a process that uses a machine-learning technique called vectorization, which identities the unique "fingerprint" of a piece of content, and finds matches along other pieces of content, as well as members working in those areas. It created a pilot where it vectorized a year's worth of content—35,000 items—and shared its findings with a select group of members. "We said we're building a recommendations engine that's AI based on natural language processing, and here's what it would recommend for you," Lurie said. "Are these pieces of content and these connections relevant to you and your research?"
Overwhelmingly, the answer was "yes."
Please select this link to read the complete article from Associations Now.