iFLYTEK and National Science Library of Chinese Academy of Sciences Collaboratively Introducing “Vertical Domain Large Model for Science and Technology Literature”

During the 10th Meeting of China Artificial Intelligence Industry Alliance (AIIA) held in Chongqing Liangjiang New Area on the afternoon of December 7, the alliance officially unveiled the AIIA Top Ten Pioneer AI Application Cases and Top Ten Potential Application Cases. “Science and Technology Literature Vertical Domain Large Model and Applications”, a joint effort by iFLYTEK and the National Science Library of Chinese Academy of Sciences, distinguished itself among over 100 shortlisted cases, earning recognition as one of the top ten pioneer application cases.

AIIA, an industry organization, was initiated in 2017 by the China Academy of Information and Communications Technology (CAICT) and other entities. AIIA has launched the solicitation for the “2023 Top 10 Pioneer AI Application Cases” to identify high-value, benchmark-worthy, and practically implemented application practices in the industry to expedite the widespread adoption of artificial intelligence.

The collaborative effort between iFLYTEK and the National Science Library of Chinese Academy of Sciences to create the science and technology literature large model offers valuable assistance in addressing the research challenges of complex and scattered information in contemporary literature. Specifically, it can help alleviate the research workload, enhance research efficiency, and reduce the overall cost of scientific research.

Tailored to the intricate demands of science and technology literature processing, the large model dedicated to this domain relies on constructing a specialized corpus and extensive learning from vast amounts of science and technology literature. The model is adept at efficient information extraction and intelligent processing, ensuring both depth and breadth of literature analysis and enhancing the efficiency of knowledge acquisition. Targeting the domain of scientific research, the science and technology literature large model endows itself with a more robust foundational capability for literature knowledge services, encompassing information extraction, paper refinement, reading comprehension, academic translation, and much more.

Spark Research Assistant, an application product built upon the science and technology literature large model, features three core functions of findings research, paper exploration, and academic writing support. It serves academic researchers, enterprise professionals, and other users, contributing to the smooth progression of scientific research.