Sign In

Communications of the ACM

ACM TechNews

Automated Coding Could Mean Big Changes for Software Developers


View as: Print Mobile App Share: Send by email Share on reddit Share on StumbleUpon Share on Hacker News Share on Tweeter Share on Facebook
circuit traces and binary digits, illustration

Credit: Getty Images

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.

From ZDNet
View Full Article

 

Abstracts Copyright © 2020 SmithBucklin, Washington, DC, USA


 

No entries found