Researchers from Intel and the Massachusetts Institute of Technology have unveiled a machine learning tool that brings developers one step closer to automated coding.
MISIM, or machine inferred code similarity technology, analyzes snippets of code to understand a piece of software's intended function, then constructs a database in which software code with similar outcomes are assigned similarity scores. MISIM can understand a new algorithm's underlying intent, and suggest to engineers alternative programming approaches, or efficiency-boosting corrections and options.
Intel's Justin Gottschlich believes MISIM could have a great effect on productivity, as software development grows in complexity. This invites speculation that the algorithm could eventually be used by any member of the general public with a good software concept.
View Full Article
Abstracts Copyright © 2020 SmithBucklin, Washington, DC, USA
No entries found