A discovery, project, tool or model is measured by the coolness of its
acronym. There are BRAIN, and MUSIC, STEM and ICARUS and the list goes on and
on (check this out). But finding the right acronym is not that easy. A short personal story:
I once worked on a quantum mechanical problem called Entanglement Sudden Death,
ESD. It's a macabre name for a really cool phenomenon. However, I showed that
entanglement, even after its sudden death, can be brought back to life. So I
really wanted a cool acronym for it and I found the acronym ESD-CPR (which
usually stands for Cardio-Respiratory Resuscitation). So I had my acronym, but
I had to find what it stood for. After hours of agony I came up with Controlled
Partial Resuscitation, which is actually a good description of the phenomenon.
Wouldn't it be nice if there was an automatic tool that, given the short
description of the research, gives you a cool, relevant acronym? AURA - AUtomatic Research Acronymizer. The problem is
obviously, NLP (Natural Language Processing), which requires to
"read" the research description, "analyze" what is written
and "come up" with the correct acronym. However, related tools are
already out there, such as the HAHAcronym.
I believe that combining tools such as the tool cited above (a database
of known acronyms and an algorithm for generating acronyms), as well as semantic
networks such as concept net,
that will accommodate the network of scientific jargon, can facilitate this
very important, highly influential and greatly required tool.
No comments:
Post a Comment