{"id":13549,"date":"2017-11-07T06:12:59","date_gmt":"2017-11-07T06:12:59","guid":{"rendered":"https:\/\/www.revoscience.com\/en\/?p=13549"},"modified":"2017-11-07T06:16:12","modified_gmt":"2017-11-07T06:16:12","slug":"artificial-intelligence-aids-materials-fabrication","status":"publish","type":"post","link":"https:\/\/www.revoscience.com\/en\/artificial-intelligence-aids-materials-fabrication\/","title":{"rendered":"Artificial intelligence aids materials fabrication"},"content":{"rendered":"

System could pore through millions of research papers to extract \u201crecipes\u201d for producing materials.<\/strong><\/em><\/span><\/p>\n

\"\"
A team of researchers at MIT, the University of Massachusetts at Amherst, and the University of California at Berkeley hope to close the materials-science automation gap, with a new artificial-intelligence system that would pore through research papers to deduce \u201crecipes\u201d for producing particular materials.
Image: Chelsea Turner\/MIT<\/figcaption><\/figure>\n

CAMBRIDGE, Mass. —\u00a0In recent years, research efforts such as the\u00a0Materials Genome Initiative<\/a>\u00a0and the\u00a0Materials Project<\/a>\u00a0have produced a wealth of computational tools for designing new materials useful for a range of applications, from energy and electronics to aeronautics and civil engineering.<\/span><\/p>\n

But developing processes for producing those materials has continued to depend on a combination of experience, intuition, and manual literature reviews.<\/span><\/p>\n

A team of researchers at MIT, the University of Massachusetts at Amherst, and the University of California at Berkeley hope to close that materials-science automation gap, with a new artificial-intelligence system that would pore through research papers to deduce \u201crecipes\u201d for producing particular materials.<\/span><\/p>\n

\u201cComputational materials scientists have made a lot of progress in the \u2018what\u2019 to make \u2014 what material to design based on desired properties,\u201d says Elsa Olivetti, the Atlantic Richfield Assistant Professor of Energy Studies in MIT\u2019s Department of Materials Science and Engineering (DMSE). \u201cBut because of that success, the bottleneck has shifted to, \u2018Okay, now how do I make it?\u2019\u201d<\/span><\/p>\n

The researchers envision a database that contains materials recipes extracted from millions of papers. Scientists and engineers could enter the name of a target material and any other criteria \u2014 precursor materials, reaction conditions, fabrication processes \u2014 and pull up suggested recipes.<\/span><\/p>\n

As a step toward realizing that vision, Olivetti and her colleagues have developed a machine-learning system that can analyze a research paper, deduce which of its paragraphs contain materials recipes, and classify the words in those paragraphs according to their roles within the recipes: names of target materials, numeric quantities, names of pieces of equipment, operating conditions, descriptive adjectives, and the like.<\/span><\/p>\n

In a\u00a0paper<\/a>\u00a0appearing in the latest issue of the journal\u00a0Chemistry of Materials<\/em>, they also demonstrate that a machine-learning system can analyze the extracted data to infer general characteristics of classes of materials \u2014 such as the different temperature ranges that their synthesis requires \u2014 or particular characteristics of individual materials \u2014 such as the different physical forms they will take when their fabrication conditions vary.<\/span><\/p>\n

Olivetti is the senior author on the paper, and she\u2019s joined by Edward Kim, an MIT graduate student in DMSE; Kevin Huang, a DMSE postdoc; Adam Saunders and Andrew McCallum, computer scientists at UMass Amherst; and Gerbrand Ceder, a Chancellor\u2019s Professor in the Department of Materials Science and Engineering at Berkeley.<\/span><\/p>\n

Filling in the gaps<\/strong><\/span><\/p>\n

The researchers trained their system using a combination of supervised and unsupervised machine-learning techniques. \u201cSupervised\u201d means that the training data fed to the system is first annotated by humans; the system tries to find correlations between the raw data and the annotations. \u201cUnsupervised\u201d means that the training data is unannotated, and the system instead learns to cluster data together according to structural similarities.<\/span><\/p>\n

Because materials-recipe extraction is a new area of research, Olivetti and her colleagues didn\u2019t have the luxury of large, annotated data sets accumulated over years by diverse teams of researchers. Instead, they had to annotate their data themselves \u2014 ultimately, about 100 papers.<\/span><\/p>\n

By machine-learning standards, that\u2019s a pretty small data set. To improve it, they used an algorithm developed at Google called Word2vec. Word2vec looks at the contexts in which words occur \u2014 the words\u2019 syntactic roles within sentences and the other words around them \u2014 and groups together words that tend to have similar contexts. So, for instance, if one paper contained the sentence \u201cWe heated the titanium tetracholoride to 500 C,\u201d and another contained the sentence \u201cThe sodium hydroxide was heated to 500 C,\u201d Word2vec would group \u201ctitanium tetracholoride\u201d and \u201csodium hydroxide\u201d together.<\/span><\/p>\n

With Word2vec, the researchers were able to greatly expand their training set, since the machine-learning system could infer that a label attached to any given word was likely to apply to other words clustered with it. Instead of 100 papers, the researchers could thus train their system on around 640,000 papers.<\/span><\/p>\n

Tip of the iceberg<\/strong><\/span><\/p>\n

To test the system\u2019s accuracy, however, they had to rely on the labeled data, since they had no criterion for evaluating its performance on the unlabeled data. In those tests, the system was able to identify with 99 percent accuracy the paragraphs that contained recipes and to label with 86 percent accuracy the words within those paragraphs.<\/span><\/p>\n

The researchers hope that further work will improve the system\u2019s accuracy, and in ongoing work they are exploring a battery of deep learning techniques that can make further generalizations about the structure of materials recipes, with the goal of automatically devising recipes for materials not considered in the existing literature.<\/span><\/p>\n

Much of Olivetti\u2019s prior research has concentrated on finding more cost-effective and environmentally responsible ways to produce useful materials, and she hopes that a database of materials recipes could abet that project.<\/span><\/p>\n","protected":false},"excerpt":{"rendered":"

System could pore through millions of research papers to extract \u201crecipes\u201d for producing materials. CAMBRIDGE, Mass. —\u00a0In recent years, research efforts such as the\u00a0Materials Genome Initiative\u00a0and the\u00a0Materials Project\u00a0have produced a wealth of computational tools for designing new materials useful for a range of applications, from energy and electronics to aeronautics and civil engineering. But developing […]<\/p>\n","protected":false},"author":6,"featured_media":13551,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[22,17,28],"tags":[],"class_list":["post-13549","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-other","category-research","category-techbiz"],"featured_image_urls":{"full":["https:\/\/www.revoscience.com\/en\/wp-content\/uploads\/2017\/11\/MIT-Materials-AI_0.gif",639,426,false],"thumbnail":["https:\/\/www.revoscience.com\/en\/wp-content\/uploads\/2017\/11\/MIT-Materials-AI_0-150x150.gif",150,150,true],"medium":["https:\/\/www.revoscience.com\/en\/wp-content\/uploads\/2017\/11\/MIT-Materials-AI_0-300x200.gif",300,200,true],"medium_large":["https:\/\/www.revoscience.com\/en\/wp-content\/uploads\/2017\/11\/MIT-Materials-AI_0.gif",639,426,false],"large":["https:\/\/www.revoscience.com\/en\/wp-content\/uploads\/2017\/11\/MIT-Materials-AI_0.gif",639,426,false],"1536x1536":["https:\/\/www.revoscience.com\/en\/wp-content\/uploads\/2017\/11\/MIT-Materials-AI_0.gif",639,426,false],"2048x2048":["https:\/\/www.revoscience.com\/en\/wp-content\/uploads\/2017\/11\/MIT-Materials-AI_0.gif",639,426,false],"ultp_layout_landscape_large":["https:\/\/www.revoscience.com\/en\/wp-content\/uploads\/2017\/11\/MIT-Materials-AI_0.gif",639,426,false],"ultp_layout_landscape":["https:\/\/www.revoscience.com\/en\/wp-content\/uploads\/2017\/11\/MIT-Materials-AI_0.gif",639,426,false],"ultp_layout_portrait":["https:\/\/www.revoscience.com\/en\/wp-content\/uploads\/2017\/11\/MIT-Materials-AI_0.gif",600,400,false],"ultp_layout_square":["https:\/\/www.revoscience.com\/en\/wp-content\/uploads\/2017\/11\/MIT-Materials-AI_0.gif",600,400,false],"newspaper-x-single-post":["https:\/\/www.revoscience.com\/en\/wp-content\/uploads\/2017\/11\/MIT-Materials-AI_0.gif",639,426,false],"newspaper-x-recent-post-big":["https:\/\/www.revoscience.com\/en\/wp-content\/uploads\/2017\/11\/MIT-Materials-AI_0.gif",540,360,false],"newspaper-x-recent-post-list-image":["https:\/\/www.revoscience.com\/en\/wp-content\/uploads\/2017\/11\/MIT-Materials-AI_0.gif",95,63,false],"web-stories-poster-portrait":["https:\/\/www.revoscience.com\/en\/wp-content\/uploads\/2017\/11\/MIT-Materials-AI_0.gif",639,426,false],"web-stories-publisher-logo":["https:\/\/www.revoscience.com\/en\/wp-content\/uploads\/2017\/11\/MIT-Materials-AI_0.gif",96,64,false],"web-stories-thumbnail":["https:\/\/www.revoscience.com\/en\/wp-content\/uploads\/2017\/11\/MIT-Materials-AI_0.gif",150,100,false]},"author_info":{"info":["Amrita Tuladhar"]},"category_info":"Other<\/a> Research<\/a> Tech<\/a>","tag_info":"Tech","comment_count":"0","_links":{"self":[{"href":"https:\/\/www.revoscience.com\/en\/wp-json\/wp\/v2\/posts\/13549","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.revoscience.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.revoscience.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.revoscience.com\/en\/wp-json\/wp\/v2\/users\/6"}],"replies":[{"embeddable":true,"href":"https:\/\/www.revoscience.com\/en\/wp-json\/wp\/v2\/comments?post=13549"}],"version-history":[{"count":0,"href":"https:\/\/www.revoscience.com\/en\/wp-json\/wp\/v2\/posts\/13549\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.revoscience.com\/en\/wp-json\/wp\/v2\/media\/13551"}],"wp:attachment":[{"href":"https:\/\/www.revoscience.com\/en\/wp-json\/wp\/v2\/media?parent=13549"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.revoscience.com\/en\/wp-json\/wp\/v2\/categories?post=13549"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.revoscience.com\/en\/wp-json\/wp\/v2\/tags?post=13549"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}