Front-end Field Binding Dataset for Code Generation

Front-end Field Binding Dataset is a dataset used to automatically generate corresponding code snippet based on natural language description, in the actual front-end development work. It includes 3 files, train.xlsx is a training file containing 2150 training samples; test.xlsx is a test file containing 150 test samples. Both training and testing samples consist of two parts: the former is the generated code, and the latter is the natural language description. variable.xlsx is a variable vocabulary, containing 15525 pieces of information，each piece is a variable name that is often used in the front end and its corresponding Chinese description. It can be used as additional information to assist code generation. The task is to automatically generate corresponding code snippet based on natural language descriptions. Front-end Field Binding Dataset is provided by F(x) Team of Taobao Department , Alibaba Group and Peking University.
<br />
<br />
Chinese：Front-end Field Binding Dataset 是一个用于评价在实际前端开发工作中，基于自然语言描述自动生成相应代码片段的数据集。其包括 3 个文件，train.xlsx 是包含 2150 条训练样本的训练文件；test.xlsx 是包含 150 条测试样本的测试文件。训练和测试样本都包含两部分：前者是生成的代码，后者是自然语言描述。variable.xlsx 是一个变量词表，包含 15525 条信息，每条是前端中经常被使用到的变量名和它对应的中文描述，可以作为额外信息辅助代码生成。任务是需要基于自然语言描述，自动生成相应的代码片段。本数据集由阿里巴巴淘系技术部 F(x) Team 和北京大学高可信软件技术教育部重点实验室联合提供。

搜集汇总

背景与挑战

背景概述

该数据集是一个用于前端开发中基于自然语言描述自动生成代码片段的数据集，包含2150条训练样本、150条测试样本以及15525条变量词表作为辅助信息。它由阿里巴巴淘系技术部F(x) Team与北京大学联合提供，遵循CC BY-NC 4.0协议。