Researchers often need to go beyond Google to find the kind of medical journal articles and flat data files necessary for their work. But many journal articles are locked away in databases like JSTOR or PubMed, which don’t have the reliable search capabilities of an engine like Google — so researchers have to waste time tracking them down.
Helix uses machine learning to suggest citations and relevant research as you write. Helix scans a writer’s text as he or she types and automatically pulls in recommendations for relevant journal articles, news and Wikipedia pages. The recommendations display in a queue alongside the writer’s main text, so they can be reviewed at a glance without leaving the word processor. A writer can request suggestions on a specific phrase or sentence by highlighting it, or just let Helix make suggestions based on the entirety of the article.
“Many researchers use Google searches because journal sites have terrible search functions,” Burke says. He and Krishnan developed Helix to give researchers a faster option that won’t distract them or take them away from their writing.
Burke and Krishnan primarily focused on medical research for their work at the Disrupt NY Hackathon, pulling journal articles from PubMed, but the duo hope to expand by adding other databases of journals as well.
Burke and Krishnan built Helix using a free trial version of Lateral and IBM Watson. Unfortunately, when their free trial expires in two weeks, Helix might face some hiccups. But by then, Burke and Krishnan hope to approach Lateral about working together.
Helix came together at the very last moment, as Hackathon projects often do. “The cutoff was 9:30 and I wrote the last line of code at 9:25,” Burke says with a laugh. “Before that, it wasn’t showable.” By the time of the demo, Helix looked impressive. Check it out below.