All of science involves research but finding just the right information can be a daunting task when it means having to comb through hundreds of sources from scientific journals. For any scientist or research team it’s a challenge to find data that is on point for a particular study or field. This is especially true in neuroscience. The expression, “Oh come on, it’s not brain surgery” doesn’t apply, because, well, sometimes it actually is. There are online databases of just about everything and in a way that is the problem. Currently there are about 34,000 peer reviewed scholarly journals in print and online. Collectively that amounts to about 2.5 million articles published each year. That’s a lot to look at for a researcher who needs information on, say, the cerebral cortex and audio processing.
Artificial intelligence (AI) is fast becoming a tool for scientists trying to sort through so much information. Computer programs that are engineered to actually analyze data for it’s relevance are leading the way over basic databases that can only return results that match specific keywords or other input. Artificial intelligence technology was developed in part with the help of neuroscience professionals so it seems only reasonable that an AI product was recently used to rank the top ten most influential neuroscience researchers. While it might seem that such a list would be difficult to develop, an AI search engine got some decent results recently.
Semantic Scholar, a program developed by the AI non-profit Allen Institute for Artificial Intelligence (AI2) looked at neuroscience research all over the globe. Universities, labs and government agencies were all part of the query to find the researchers (and their work) that had most influenced neuroscience. A daunting task indeed, but it wasn’t the first time the program had been used to rank professionals who collectively had contributed millions of pages and hundreds of thousands of hours of work to their field. In April, the program was used to rank the most influential computer scientists. Those results were based on looking at over two million papers in the field of CS and sorting them intelligently based on a number of factors.
The recent ranking of the most influential neuroscientists was an even larger project, drawing the results from a corpus of 10 million research papers, 25% of which concerned neuroscience topics. CEO of AI2 Oren Etizioni told Science Magazine, “We are using machine learning, natural language processing, and [machine] vision to begin to delve into the semantics. Think Siri for science, only better”
While having a tool that can parse actual data into intelligent results seems like something only technology geeks would care about, it’s actually a significant advance for medical research. In a press release AI2 founder Paul Allen said, “Medical breakthroughs should not be hindered by the cumbersome process of searching the scientific literature. My vision for Semantic Scholar is to give researchers more powerful tools to comb through millions of academic papers online, to help them keep up with the explosive growth of science.”
AI2 hopes to expand it’s database of research. There are plans to build the corpus to include all biomedical research (as defined by PubMed) by the end of 2017. Check out the video to see how all of this information can advance medical science
Sources Science Magazine Tech Crunch Discover Magazin