A high imbalance exists between technical debt and non-technical debt source code comments. Such imbalance affects Self Admitted Technical Debt (SATD) detection performance, and existing literature lacks empirical evidence on the choice of balancing technique. In this work, we evaluate the impact of multiple balancing techniques, including Data level, Classifier level, and Hybrid, for SATD detection in Within-Project and Cross-Project setup. Our results show that the Data level balancing technique SMOTE or Classifier level Ensemble approaches with Random Forest or XGBoost are reasonable choices depending on whether the goal is to maximize Precision, Recall, F1, or AUC-ROC. We compared our best-performing model with the previous SATD detection benchmark (cost-sensitive Convolution Neural Network). Interestingly the top-performing XGBoost with SMOTE sampling improved the Within-project F1 score by 10% but fell short in Cross-Project set up by 9%. This supports the higher generalization capability of deep learning in Crossproject SATD detection, yet while working within individual projects, classical machine learning algorithms can deliver better performance. We also evaluate and quantify the impact of duplicate source code comments in SATD detection performance. Finally, we employ SHAP and discuss the interpreted SATD features. We have included the replication package 1 and shared a web-based SATD prediction tool 2 with the balancing techniques in this study.
With the proliferation and ubiquity of smart gadgets and smart devices, across the world, data generated by them has been growing at exponential rates, in particular social media platforms like Facebook, Twitter and Instagram have been generating voluminous data on a daily basis. According to Twitter's usage statistics, about 500 million tweets are generated each day. While the tweets reflect the users' opinions on several events across the world, there are tweets which are offensive in nature that need to be tagged under the hateful conduct policy of Twitter. Offensive tweets have to be identified, captured and processed further, for a variety of reasons, which include i) identifying offensive tweets in order to prevent violent/abusive behaviour in Twitter (or any social media for that matter), ii) creating and maintaining a history of offensive tweets for individual users (would be helpful in creating meta-data for user profile), iii) inferring the sentiment of the users on particular event/issue/topic. We (CodaLab Team/User Name: murali sr) have employed neural network models which manipulate attention with Temporal Convolutional Neural Network for the three shared subtasks i) ATT-TCN (ATTention based Temporal Convolutional Neural Network) employed for shared sub-task A that yielded a best macro-F1 score of 0.46, ii) SAE-ATT-TCN(Self Attentive Embedding-ATTention based Temporal Convolutional Neural Network) employed for shared sub-task B and sub-task C that yielded best macro-F1 score of 0.61 and 0.51 respectively. Among the two variants ATT-TCN and SAE-ATT-TCN, the latter performed better.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.