Home health remedies Elsevier research partner Karin Verspoor nominated For Women in AI award

Elsevier research partner Karin Verspoor nominated For Women in AI award


Posted on April 5th, 2022 by in AI & Data


In recognition of her numerous research accomplishments, Professor Karin Verspoor is nominated for the prestigious Women in AI Award under the ‘AI in Innovation’ category. Certainly, the interdisciplinary work she has undertaken with Elsevier has chemists and pharmaceutical companies excited about the real-world impact of automating the extraction of chemical reaction information from patents.

Getting chemists excited

Professor Karin Verspoor is an Elsevier collaborator and world-renowned NLP (natural language processing) specialist – and now she is also a finalist for the Women in AI (WAI) Award. Over the past four years, Elsevier Life Sciences has enjoyed a very productive research collaboration with Karin and her research team in Australia via our ChEMU (Cheminformatics Elsevier Melbourne Universities) project. 

The ChEMU project represents the first deep dive into automating the process of extracting information about chemical reactions in chemical patents – not just from texts but also from tables. “This collaboration has already resulted in lots of data sets, twenty research publications and two patents, with lots more to follow as a result of our community challenges,” enthuses Dr. Saber Akhondi, Director at Data Science Life Sciences Elsevier, who leads the team composed of Dr. Camilo Thorne and Dr. Christian Druckenbrodt on the Elsevier side. 

With a PhD based on extracting chemical information from patents, Saber is very aware of the groundbreaking nature of Karin’s work. “What originally impressed me most about Karin was how she went beyond her specialty of NLP and really sat down with us to dig deep and understand what the problems of chemists are and what gets them excited.”

Celebrating women in AI

The 2022 Women in AI Awards honor and celebrate the achievements of Australian and New Zealand women in the field of AI across 14 categories, with this year’s theme being “The Transformative Nature of AI”. 

“Finalists are judged on innovation, leadership and inspiring potential, global potential and impact, and ability for the AI solution to do good for the community and citizens,” according to the WAI organization. 

Empowering Pharma to work better and faster

As Dean of the School of Computing Technologies at RMIT University in Melbourne, Australia, Karin ticks all the WAI boxes. Her research mainly focuses on the use of AI to spur biological discovery and clinical decision support via the extraction of information from clinical texts and medical literature.

“I’ve been a woman in AI for nearly 30 years,” says Karin, “Starting when I was an undergraduate computer science student. I was drawn to AI because I was fascinated by our human ability to understand language, and I saw AI as a way for us to gain insight into how language works through computational models.”

Her ChEMU collaboration is aimed at helping chemists and pharma companies do their work as efficiently and effectively as possible. Essentially, the project is out to further activate the chemical information stored in Elsevier’s Reaxys database of 170 million chemical structures, reactions and properties derived from over 95 million patent literature and journal articles. 

“The AI technologies we are developing facilitate rapid ingestion and processing of these documents and make the valuable information contained within them structured and searchable,” says Karin. “This has broad implications for scientific hypothesis generation, effective drug development, and beyond.”

The challenges of patents

“What makes patents so interesting is that they are the place where information comes out first,” says Saber. “Since it usually takes four to six years for data in a patent to be published in a journal.” 

But while the patent information may be the ‘freshest’, that does not mean it’s easy to access. “With patents, and especially in the chemistry domain, there’s a lot of focus on trying to hide the information – so it’s not immediately obvious to potential competitors,” notes Saber. 

There is also a whole array of other challenges: from varying data formats to the intrinsic complexity of a reaction process. “And all this information is not placed in a single paragraph – it’s spread across the patent and across texts, illustrations and tables,” says Saber. “So, you can say ChEMU is a project that brings together many different projects.” 

Sharing the love

Another key element of ChEMU is running annual community challenges based on opening up datasets to stimulate further research related to, for example, entity recognition or event extraction. “We made various very important and unique data sets open to both industry and academia for research purposes,” says Saber.

“Currently these teams are all beginning to publish, and this works to highlight the effectiveness of different methodologies in addressing the same problem. This not only helps Elsevier to take its technology forward, but also allows the greater community to go forward in this area and create more interest.”

A true collaboration

Funded equally by Elsevier and the Australian government, ChEMU involves two PhD students and several postdocs. “It wasn’t just two groups working together,” says Saber. “It was really a collaborative atmosphere, with a lot of back-and-forth, in term of both chemistry and data science. I believe the students were very stimulated by the fact their work had a commercial focus.”

“We need each other – it’s absolutely a collaboration,” says Karin. “Sometimes academia and companies are on different time scales, but I think we all feel like we’re pulling in the same direction on this, and it’s just a real pleasure to see that level of commitment from both sides.”

Transforming healthcare through applied research

“What’s remarkable about this collaboration is that the deeper we went, the more opportunities we saw. And in the process, we won over the hearts and minds of many chemists,” says Saber. “They get excited when they see us, for example, expand our reach from seven, first to twelve, then up to a hundred patent offices. This really shows them the power of AI: that the more AI can get involved, the less time and costs will be needed to bring new solutions to market.”

As the original graduate students round off their work, this phase of the collaboration is coming to an end. “We are now talking about how we can continue working together. Certainly, we want to continue with the community challenges,” says Saber. 

As Karin puts it: “I’m pleased that my work in this area has been recognized. Interdisciplinary work such as this sometimes is less visible than the big methodological innovations, but it is definitely no less important.”

Read: ‘How data scientists are uncovering chemical compounds hidden in patents

Please enable JavaScript to view the comments powered by Disqus.

R&D Solutions for Pharma & Life Sciences

We’re happy to discuss your needs and show you how Elsevier’s Solution can help.

Contact Sales

{n.callMethod? n.callMethod.apply(n,arguments):n.queue.push(arguments)}
fbq(‘init’, ‘533182150132648’);
fbq(‘track’, “PageView”);
{n.callMethod? n.callMethod.apply(n,arguments):n.queue.push(arguments)}
fbq(‘init’, ‘1737613393127776’,
{ em: ‘insert_email_variable,’ }
fbq(‘track’, ‘PageView’);

Source link


Please enter your comment!
Please enter your name here

sixteen − 12 =