Imagine millions of scientific papers containing valuable experimental data, sitting like unopened treasure chests waiting to be discovered. Thanks to an inspiring collaboration between researchers and artificial intelligence, these hidden gems are finally being brought to light.

Dr. Yukari Katsura, a Senior Researcher at Japan's National Institute for Materials Science, is leading a remarkable initiative that's transforming how scientists share and use experimental data. Her team has developed two groundbreaking tools that harness the power of large language models—the same technology behind ChatGPT—to extract precious materials science data from research papers published over decades.

"Graphs in the millions of papers published to date contain valuable experimental data collected by past researchers, and much of it remains untapped," explains Dr. Katsura, whose passion for democratizing scientific knowledge shines through her work.

The story begins in 2015 when Dr. Katsura launched Starrydata, an ambitious project to build a comprehensive materials property database. Initially, dedicated researchers painstakingly collected data manually, successfully amassing an unprecedented volume of experimental information. Now, with artificial intelligence as their partner, the process is becoming faster and more efficient than ever.

The first tool, Starrydata Auto-Suggestion, acts like a helpful research assistant. When scientists paste text from a paper, it automatically suggests relevant data entries, saving countless hours of work. The second tool, Starrydata Auto-Summary, takes an entire research paper and organizes all the experimental data into an easy-to-read format, helping researchers quickly find exactly what they need.

What makes this work particularly exciting is its potential impact on everyday life. The functional materials that scientists are studying form the backbone of modern technology—from the smartphones in our pockets to the electric vehicles driving us toward a sustainable future. By making experimental data more accessible, researchers worldwide can accelerate the development of even better materials.

The beauty of this approach lies in its collaborative spirit. Rather than replacing human expertise, the AI tools enhance what researchers can accomplish. While the technology handles time-consuming extraction tasks, human scientists apply their knowledge and creativity to interpret the data and drive innovation forward.

Dr. Katsura envisions an inspiring future where experimental data from all materials science fields can be shared digitally and viewed comprehensively. "By deconstructing papers and returning them to the form of experimental data, other researchers can also use it for their own research," she explains.

The project is already making waves internationally. Leading researchers worldwide are beginning to utilize Starrydata's open datasets for developing new materials. While the database currently focuses on specific areas like thermoelectric materials and magnets, the team's ambitious vision extends to encompassing all materials science fields.

This heartwarming story of human-AI collaboration demonstrates how technology can amplify our collective knowledge and accelerate scientific progress. By unlocking data that has been buried in papers for years, Dr. Katsura and her team are building bridges between past discoveries and future innovations, creating opportunities for breakthroughs that could benefit us all.