Semantic scholars at the Allen Institute can now search 175 million academic papers.

2019-10-24

Some studies have shown that over 3 million scientific papers are published in English every year, with thousands being published daily. Apart from the authors and journal editors, only a fraction of these papers can be read by anyone.


To assist the vast number of papers, the Allen Institute for AI, founded by the late Microsoft co-founder Paul Allen, released Semantic Scholar in 2015. This public search engine technology combines machine learning, natural language processing, and computer vision to highlight data from computer science and biomedical journal papers and determine their connections. To date, over 2 million users have adopted Semantic Scholar to analyze academic literature, covering topics from gender bias in clinical research to the accelerated development of AI research in China. Now, the Allen Institute aims to establish a semantic foundation for the next millions of users.


Prior to this week, Semantic Scholar covered a total of 40 million research papers, along with related blog posts, news articles, videos, and other resources. However, it has now more than doubled its coverage to over 175 million papers in all scientific fields, including natural sciences such as biology, chemistry, geology, materials science, and medical physics. According to the Allen Institute, this growth makes Semantic Scholar the most comprehensive search engine for academic content worldwide.


Doen Raymond, General Manager of the Allen Institute, said, "Search engines cannot be used for scientific research in the same way as they are used for shopping, finding restaurants, or searching for news articles. Successful scientific search must utilize AI to understand scientific papers and enable researchers to go beyond keywords to find the right information. That's the purpose behind Semantic Scholar. We are now at a crucial moment in history where every scientist can easily access a powerful and free AI search engine."Raymond, a retired US Army veteran, developed machine learning programs for Amazon's Alexa platform and Google's search monetization division in the Asia-Pacific region. He was hired to lead the Semantic Scholar project in March 2018. In December, the Allen Institute announced its collaboration with Microsoft to connect Semantic Scholar with the latter's "Academic Graph," a heterogeneous graph containing scientific publication records and citation relationships, as well as authors, institutions, journals, conferences, and research fields.


With the addition of the Academic Graph records, the number of papers indexed in Semantic Scholar increased to over 175 million by August 2019. They also indirectly support the development of Supp AI.


Oren Etzioni, CEO of the Allen Institute, said, "All universities and scientists today should be studying how AI and deep learning can advance science. We have demonstrated that Semantic Scholar not only enables deep semantic search but also allows for in-depth exploration of published papers in any scientific field."

RECENT AI NEWS

RECENT AI TOOLS