Federated learning based multi‐task feature fusion framework for code expressive semantic extraction. (16th May 2022)

• Record Type: Journal Article
• Authors: Deng, Fengyang; Fu, Cai; Qian, Yekui; Yang, Jia; He, Shuai; Xu, Hao
• Is Part Of: Software, practice & experience. Volume 52:Number 8(2022)
• Page Start: 1849
• Page End: 1866
• Publication Date: 2022-05-16
• Subjects: code completion -- cross‐language code analysis -- feature fusion -- federated learning -- graph neural network -- multi‐task -- variable misuse; Computer software -- Periodicals; Computer programming -- Periodicals; Computer programs -- Periodicals; 005.3
• DOI: 10.1002/spe.3094
• Languages: English
• ISSNs: 0038-0644

Abstract

Abstract: Using multi‐task learning to extract code features can effectively increase the information of the features. However, the existing multi‐task learning methods mainly have two limitations: (1) They cannot extract enough code‐related information or only extract similar semantic features. Similar multi‐task makes the information in the features increased insufficiently. However, the high difference multi‐task is challenging to converge. (2) They cannot train multi‐task on heterogeneous datasets. In standard multi‐task training, we need to label all tasks for all data, which consumes enormous labor. To solve the above limitations, we select two high difference tasks, the cross‐language code completion task and variable misuse task, to extract expressive semantic code features. We propose an attention‐based feature fusion module to merge information among high difference tasks, avoiding the convergence dilemma of standard multi‐task learning. We propose a federated learning framework, extracting semantic information and using the feature fusion module to integrate multi‐task information among single labeled datasets. We experiment on C# and Python datasets for the code completion and variable misuse tasks. The results show that the performance of fusion features by FedMTFF improved by up to 22.6% and 15.1% compared to single tasks. We use FedMTFF to perform four cross‐language multi‐task features fusion, exceeding the current best baseline by 24.1%.