The Challenge of Managing Data Science Change (VIDEO)

Video produced by Steve Nathans-Kelly

Elsevier technology research director Helena Deus discussed the challenges Elsevier has faced in its efforts to standardize quality metrics and improve curation across the enterprise in this clip from her keynote at Data Summit 2019.

In the presentation, titled “Digital Transformation Is Business Transformation: How to Incorporate AI Technology Into a 130-Year-Old Company,” Deus showcased how the company combines content and data with analytics and technology to help researchers to make new discoveries and have more impact on society, and clinicians to treat patients better and save more lives.

Deus was presenting the work by Dr. Michelle Gregory, SVP Data Science, Elsevier.

Here's the challenge we face and we're still facing it on a daily basis: Before we made this transition to analytics, we used to hire contractors to extract information from text. It was a brute-force approach, but you should never underestimate the value of brute force. So we would hire them to process the information in the back end and they would then create structured information to power the tools to help the decision-making process,” explained Deus.

DBTA’s next Data Summit conference will be held May 19-20, 2020, in Boston, with pre-conference workshops on Monday, May 18.

However, said Deus, “for each product, because we had to hire experts from many different fields, there were chemists, biologists, doctors ... each product had its own quality metrics, each product has its own tools and content enrichment tools. And so this was quite a bit of a challenge because we wanted the ability to have a single quality metric across the organization.”

In some cases, said Deus, customers were very unhappy about the amount of content they had for the decision-making process while contractors were saying, "Oh, but we have 99% accuracy; what are you saying, the customers are complaining?"

According to Deus, “It turned out it was the way that our contractors were thinking about quality metrics that was not shared across the organization. Beyond that, as we were in this process, we were not able to iterate, we were not able to check: What is the best workflow; what is the best way to actually get structured content out of text; and to put it on the tools. We couldn't incorporate user behavior, there was not a lot of automation we could do. We couldn't improve the curation process because it was all manual and devised by the contractors themselves.

Many presenters have made their slide decks available on the Data Summit 2019 website at

To access the full keynote, titled “Digital Transformation Is Business Information: How to Incorporate AI Technology into a 130-Year-Old Company,” go to