Bimodal Modelling of Source Code and Natural Language
نویسندگان
چکیده
We consider the problem of building probabilistic models that jointly model short natural language utterances and source code snippets. The aim is to bring together recent work on statistical modelling of source code and work on bimodal models of images and natural language. The resulting models are useful for a variety of tasks that involve natural language and source code. We demonstrate their performance on two retrieval tasks: retrieving source code snippets given a natural language query, and retrieving natural language descriptions given a source code query (i.e., source code captioning). Experiments show there to be promise in this direction, and that modelling the structure of source code improves performance.
منابع مشابه
AutoClass: Automatic Text to OOP Concept Identification Model
Requirement specification is one of the most crucial steps in Software Development Life Cycle (SDLC). Well-written requirement documents include expressive keywords such as names of classes, attributes and methods to be used in software design. Knowledge extraction from these documents may be helpful in order to design conceptual model for Object Oriented (OO) systems. Recently, automatic text-...
متن کاملLanguage choice in bimodal bilingual development
Bilingual children develop sensitivity to the language used by their interlocutors at an early age, reflected in differential use of each language by the child depending on their interlocutor. Factors such as discourse context and relative language dominance in the community may mediate the degree of language differentiation in preschool age children. Bimodal bilingual children, acquiring both ...
متن کاملA two dimensional Simulation of crack propagation using Adaptive Finite Element Analysis
Finite element method (FEM) is one of the most famous methods which has many applications in varies studies such as the study of crack propagation in engineering structures. However, unless extremely fine meshes are employed, problem arises in accurately modelling the singular stress field in the singular element area around the crack tip. In the present study, the crack growth simulation has b...
متن کاملLearning Unified Features from Natural and Programming Languages for Locating Buggy Source Code
Bug reports provide an effective way for end-users to disclose potential bugs hidden in a software system, while automatically locating the potential buggy source code according to a bug report remains a great challenge in software maintenance. Many previous studies treated the source code as natural language by representing both the bug report and source code based on bag-of-words feature repr...
متن کاملDeSoCoRe: Detecting Source Code Re-Use across Programming Languages
Source code re-use has become an important problem in academia. The amount of code available makes necessary to develop systems supporting education that could address the problem of detection of source code re-use. We present the DeSoCoRe tool based on techniques of Natural Language Processing (NLP) applied to detect source code re-use. DeSoCoRe compares two source codes at the level of method...
متن کامل