Machine Learning
The researchers from Lawrence Berkeley National Laboratory used an algorithm ‘Word2Vec’ to sift through a countless number of research papers to find ‘hidden discoveries’ that were elusive to human researchers. The said algorithm didn’t understand the meaning of a scientific term like ‘Thermoelectric materials’ and its attribute is to convert heat to energy which is used in numerous heating and cooling applications. The machine was not trained in material science to understand the nuances of scientific term. It simply compared several ‘similarly titled’ papers on materials science and could make connections that no human scientist could do, according to the Indian researcher Anubhav Jain. It also makes suggestions of cross-discipline associations’ based on the ‘word similarity.’
To train the algorithm, the researchers assessed the language in 3.3 million abstracts related to material science, ending up with a vocabulary of about 500,000 words. They fed the abstracts to Word2vec, which used machine learning to analyze relationships between words. And the result is simply unbelievable.
In an exciting experiment, researchers analyzed only papers published before 2009 and were able to predict one of the best modern-day thermoelectric materials four years before it was discovered in 2012. With the advent of ‘machine learning,’ kind of tech, the future looks better and brighter.
Compiled by
Srini
Reference Sites