AI Case Study

Rice University researchers are developing an application that writes Java API code snippets using deep learning

Rice University's Intelligent Software Systems Laboratory have created an application that can write portions of code by querying a deep learning based application called Bayou. Bayou helps decide which API data type or API method call to use. It is funded by Defense Advanced Research Projects Agency.

Industry

Public And Social Sector

Education And Academia

Project Overview

"Bayou uses neural sketch learning to train the AI neural network to notice high-level patterns among many Java programs. The associative learning process relates unique sketches for each program with the user intent behind the program.

BAYOU is an deep learning tool that basically works like a search engine for coding: tell it what sort of program you want to create with a couple of keywords, and it will spit out java code that will do what you’re looking for, based on its best guess.

Using its knowledge and sketches, Bayou takes a guess as what a user needs based on the question entered. "

Reported Results

Research; Results not yet available.

Technology

"Bayou is a system for automatic generation of API idioms in Java, powered by method called Neural Sketch Learning."
"a particular kind of probabilistic encoder-decoder, called a Gaussian Encoder-Decoder or GED, to learn a distribution over sketches conditioned on labels."

Function

R And D

Core Research And Development

Background

"Rice University researchers is developing a deep learning based model to predict the entire body of a method given just a few API calls or data types that appear in the method."

Benefits

Data

"Evaluate BAYOU in the generation of API-manipulating Android methods, using a corpus of about 150,000 methods drawn from
an online repository."