Articles, Blog

Research Screener | Curtinnovation Awards 2019


Systematic reviews and meta-analyses are considered the highest form of research evidence, involving an exhaustive search to identify
all relevant articles on a topic. However, the process of screening
thousands of abstracts is typically undertaken manually by a
panel of two researchers, often taking over 100 hours to complete. A Curtin team has developed a web app
called Research Screener that can analyse thousands of academic papers in a few
hours with a very high degree of accuracy. Their innovation uses sophisticated AI technologies including machine learning and
natural language processing, to identify and rank relevant abstracts, generating
a shortlist that is further refined by users. With the help of the app, research teams can achieve savings of up to 95% of the screening time. There is currently no commercially available
system harnessing machine learning like this, with the closest rivals unable to realise
the significant time reductions of this app. Research Screener has proven to be a
valuable tool in the research space, but also has broader potential to
support knowledge workers in legal, engineering, and management consulting disciplines. With two and a half million new
articles published every year, a system to help screen articles is needed now, and will be essential for research in the future.

Leave a Reply

Your email address will not be published. Required fields are marked *