Innovative AI tools in scientific research
In today’s rapidly evolving world, AI is not just transforming industries but also unlocking new opportunities for innovation across almost all sectors, including research. One such opportunity in scientific research is to create an objective method for evaluating the originality of research: an AI-powered tool for novelty scores. By assessing the novelty of studies, i.e. how surprising and innovative yet sensible these studies are, these scores enable industries such as pharmaceutical research & development to quickly identify and gauge areas for innovation more efficiently. This helps significantly speed up the traditional peer review processes. Professor Philipp Koellinger, Co-Founder and CEO of the preprint start-up DeSci Labs, explores how embracing AI in scientific publishing can accelerate progress and drive meaningful advancements across multiple industries.
The Novelty Scores Tool can make it easier and quicker to filter and find research that is novel and innovative—and, therefore, more likely to make a difference.
The traditional peer review process of many journals often prioritises articles that are viewed as novel or surprising. However, the process is not only frustratingly slow but also highly subjective and prone to biases. An objective measure can ensure that research is novel, unlocking the potential to drive meaningful advancements across the many sectors that rely on science.
AI is a powerful tool to help streamline processes, making them more efficient and effective. By automating tasks and providing advanced analytical capabilities, AI can significantly reduce publication times, enabling faster dissemination of research findings.
Additionally, AI can enhance the quality of research by offering deeper insights and identifying novel patterns that may otherwise go unnoticed. More generally, it can enhance the accessibility of research, potentially driving faster scientific progress and fostering innovation.
Novelty scores and other AI tools can play a role in encouraging scientists to take higher risks in the research questions they tackle, to think outside the box, and to do unusual things that can lead to breakthroughs. These breakthroughs for science would also help impact industries, many of which are in high need of innovation to solve pertinent global issues. Such industries include pharmaceuticals, energy, agriculture, and transportation – all of which heavily rely on science to propel their projects.
How does the novelty score tool work?
The Novelty Scores Tool was created to measure the originality of scientific research by evaluating two distinct aspects of scientifically published texts: content and context. Content novelty is determined by analysing the unique combinations of topics and concepts within a manuscript, while context novelty assesses the distinctiveness of the combination of topics and specialisations of articles cited in the reference list. Both scores take into account not only how surprising the combination of various fields in a publication is, but also how likely it is to be successful.
These scores are derived using a mathematical framework introduced by Professors James Evans and Feng Shi from the University of Chicago, which has shown a strong correlation between higher novelty scores and increased citation rates, even in prestigious scientific journals. By providing an objective measure of originality, novelty scores help identify research with the potential for significant impact.
How can AI tools support different industries?
This AI tool, along with others in the future, can support industries by providing broader scientific insights at high speed and enhancing reproducibility and novelty in scientific research, which are critical for driving innovation and technological progress.
For the novelty scoring system specifically, impactful and groundbreaking research can be identified in seconds, helping industries prioritise advancements with the greatest potential to lead to new products, services or solutions. It also fosters a stronger feedback loop between science and technology, where advancements in one drive progress in the other, accelerating innovation in fields like healthcare, biotechnology and beyond. Novelty scores, in this instance, demonstrate if there is room for an invention or new development, and identify when researchers and developers might be ‘reinventing the wheel’.
This not only saves time and money but helps industries harness scientific discoveries more effectively to address urgent challenges and drive meaningful progress. The traditional peer review process often takes months or even years to complete and given the subjective judgment, it makes it challenging even for experienced scientists to assess novelty accurately. This makes the process of feedback and assessment incredibly slow, and the system of scientific research less agile in responding and slow to change direction.
Moving toward more efficient and objective systems and utilising AI is essential to advancing science and driving industrial projects forward in an efficient and impactful manner, without losing innovation and creativity.
How can AI improve science?
Using an AI tool such as novelty scores can support scientists’ chances of doing research that is of high interest to publishers. Having an objective quality score that indicates a study is original, is likely to help streamline the peer review process. With the novelty already validated through advanced algorithms, there is less room for debate during review, ultimately reducing the time spent on discussions, and improving the efficiency of the process.
Additionally, researchers can use this tool to enhance the novelty of their work, as it allows researchers to check and refine their research. This ensures that the science being published not only demonstrates high levels of novelty but also meets the rigorous standards expected in scientific communities, ultimately contributing to the advancement of knowledge in the field.
You can find out more about how novelty score tools are developed, and the effect they can have on science and access your novelty score here.
Get in touch
For event sponsorship enquiries, please get in touch at oliver.toke@31media.co.uk or calum.budge@31media.co.uk
For media enquiries, please get in touch with vaishnavi.nashte@31media.co.uk
