Meta Scientific Discovery Beyond Search CHAN ZUCKERBERG INITIATIVE Alex Wade @alexwade
2
Supporting science & technology that will make it possible to cure, prevent, and manage all diseases by the end of the century. 3
CHAN ZUCKERBERG SCIENCE Supporting science and technology that will make it possible to cure, prevent, or manage all diseases by the end of the century 83 years Accelerating biomedical science by developing new tools and technologies and supporting open, collaborative models of research 10 years TITLE HERE 4
Growing Scale of Science 400x since 1950 1900 1950 2010 5
> 4000 papers published each day in biomedicine Nature Publishing Group Author Insights 2013 (n = 7,929 researchers) Nature Publishing Group Survey 2015 (n = 4,000 researchers) 6
85% of scientists believe they are missing important new research Nature Publishing Group Author Insights 2013 (n = 7,929 researchers) Nature Publishing Group Survey 2015 (n = 4,000 researchers) 7
73% of scientists wish they could be more up-to-date with the latest research in their specialty Nature Publishing Group Author Insights 2013 (n = 7,929 researchers) Nature Publishing Group Survey 2015 (n = 4,000 researchers) 8
Some Challenges to Researchers: 1. Stay up to date with the research (Narrow vs Broad) 2. Understand context and relations 3. Insight from emerging concepts
1 0
1 1
Meta is a tool to discover biomedical research in real-time Never miss an important paper in your personal feeds. See advances and emerging trends at the intersections of fields. Save key papers to read later and sync with your reference manager. 12
Meta uses artificial intelligence to analyze and connect insights across millions of papers. It seeks out the most relevant or impactful studies in a scientific area the moment they are published, and finds patterns in the literature on a scale that no human being could accomplish alone. 13
Some Challenges to Researchers: 1. Stay up to date with the research (Narrow vs Broad) 2. Understand context and relations 3. Insight from emerging concepts Feeds Predictive Ranking Recommendations Impact Ranking Trend Analysis Predictive Analytics
Meta: Product Walkthrough 16
1 7
18
Title Title
Title Title
Title 21
Title 22
Title 23
Title 24
Title 25
Title 26
Title 27
Meta s Product Focused Machine Learning Program Goals Highly Accurate Entity Graph Concept Recognition & Disambiguation Article Impact Predictions Entity Emergence Entity + article recommendations Real world collaboration graph Feed + search personalization 28
Community Collaboration Meta has been built in collaboration with biomedical institutions in the bay area. Engaged 1000+ academic biomedical researchers for UX research, qualitative and quantitative feedback on data quality and product experience. Public beta for all researchers targeted for Fall 2018. 29
Meta s Publisher Program Over 2,000 publisher partners >80M full-text peer-reviewed articles Broad text mining / information extraction rights, in exchange for unique bibliometric insights and article readership 30
31
32
33
Thank you.
Choosing Projects THE CHAN ZUCKERBERG SCIENCE ADVISORY BOARD ART LEVINSON Calico HUDA ZOGHBI Baylor College of Medicine BOB TJIAN UC Berkeley DAVID HAUSSLER UC Santa Cruz SHIRLEY TILGHMAN Princeton TOBIAS BONHOEFFER Max Planck Institute HAROLD VARMUS Weill-Cornell Medical School TITLE HERE 3 5