An official website of the United States government
Software traceability establishes a network of connections between diverse artifacts such as requirements, design, and code. However, given the cost and effort of creating and maintaining trace links manually, researchers have proposed automated approaches using information retrieval techniques. Current approaches focus almost entirely upon generating links between pairs of artifacts and have not leveraged the broader network of interconnected artifacts. In this paper we investigate the use of intermediate artifacts to enhance the accuracy of the generated trace links - focusing on paths consisting of source, target, and intermediate artifacts. We propose and evaluate combinations of techniques for computing semantic similarity, scaling scores across multiple paths, and aggregating results from multiple paths. We report results from five projects, including one large industrial project. We find that leveraging intermediate artifacts improves the accuracy of end-to-end trace retrieval across all datasets and accuracy metrics. After further analysis, we discover that leveraging intermediate artifacts is only helpful when a project's artifacts share a common vocabulary, which tends to occur in refinement and decomposition hierarchies of artifacts. Given our hybrid approach that integrates both direct and transitive links, we observed little to no loss of accuracy when intermediate artifacts lacked a shared vocabulary with source or target artifacts.
more »
« less
Traceability Transformed: Generating More Accurate Links with Pre-Trained BERT Models
Software traceability establishes and leverages associations between diverse development artifacts. Researchers have proposed the use of deep learning trace models to link natural language artifacts, such as requirements and issue descriptions, to source code; however, their effectiveness has been restricted by availability of labeled data and efficiency at runtime. In this study, we propose a novel framework called Trace BERT (T-BERT) to generate trace links between source code and natural language artifacts. To address data sparsity, we leverage a three-step training strategy to enable trace models to transfer knowledge from a closely related Software Engineering challenge, which has a rich dataset, to produce trace links with much higher accuracy than has previously been achieved. We then apply the T-BERT framework to recover links between issues and commits in Open Source Projects. We comparatively evaluated accuracy and efficiency of three BERT architectures. Results show that a Single-BERT architecture generated the most accurate links, while a Siamese-BERT architecture produced comparable results with significantly less execution time. Furthermore, by learning and transferring knowledge, all three models in the framework outperform classical IR trace models. On the three evaluated real-word OSS projects, the best T-BERT stably outperformed the VSM model with average improvements of 60.31% measured using Mean Average Precision (MAP). RNN severely underperformed on these projects due to insufficient training data, while T-BERT overcame this problem by using pretrained language models and transfer learning.
more »
« less
- Award ID(s):
- 1901059
- PAR ID:
- 10262158
- Date Published:
- Journal Name:
- Proceedings of the 43rd International Conference on Sofware Engineering
- Volume:
- 43
- Page Range / eLocation ID:
- 324 to 335
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
null (Ed.)Detecting security vulnerabilities in software before they are exploited has been a challenging problem for decades. Traditional code analysis methods have been proposed, but are often ineffective and inefficient. In this work, we model software vulnerability detection as a natural language processing (NLP) problem with source code treated as texts, and address the auto-mated software venerability detection with recent advanced deep learning NLP models assisted by transfer learning on written English. For training and testing, we have preprocessed the NIST NVD/SARD databases and built a dataset of over 100,000 files in C programming language with 123 types of vulnerabilities. The extensive experiments generate the best performance of over 93% accuracy in detecting security vulnerabilities.more » « less
-
The commonsense natural language inference (CNLI) tasks aim to select the most likely follow-up statement to a contextual description of ordinary, everyday events and facts. Current approaches to transfer learning of CNLI models across tasks require many labeled data from the new task. This paper presents a way to reduce this need for additional annotated training data from the new task by leveraging symbolic knowledge bases, such as ConceptNet. We formulate a teacher-student framework for mixed symbolic-neural reasoning, with the large-scale symbolic knowledge base serving as the teacher and a trained CNLI model as the student. This hybrid distillation process involves two steps. The first step is a symbolic reasoning process. Given a collection of unlabeled data, we use an abductive reasoning framework based on Grenander's pattern theory to create weakly labeled data. Pattern theory is an energy-based graphical probabilistic framework for reasoning among random variables with varying dependency structures. In the second step, the weakly labeled data, along with a fraction of the labeled data, is used to transfer-learn the CNLI model into the new task. The goal is to reduce the fraction of labeled data required. We demonstrate the efficacy of our approach by using three publicly available datasets (OpenBookQA, SWAG, and HellaSWAG) and evaluating three CNLI models (BERT, LSTM, and ESIM) that represent different tasks. We show that, on average, we achieve 63% of the top performance of a fully supervised BERT model with no labeled data. With only 1000 labeled samples, we can improve this performance to 72%. Interestingly, without training, the teacher mechanism itself has significant inference power. The pattern theory framework achieves 32.7% accuracy on OpenBookQA, outperforming transformer-based models such as GPT (26.6%), GPT-2 (30.2%), and BERT (27.1%) by a significant margin. We demonstrate that the framework can be generalized to successfully train neural CNLI models using knowledge distillation under unsupervised and semi-supervised learning settings. Our results show that it outperforms all unsupervised and weakly supervised baselines and some early supervised approaches, while offering competitive performance with fully supervised baselines. Additionally, we show that the abductive learning framework can be adapted for other downstream tasks, such as unsupervised semantic textual similarity, unsupervised sentiment classification, and zero-shot text classification, without significant modification to the framework. Finally, user studies show that the generated interpretations enhance its explainability by providing key insights into its reasoning mechanism.more » « less
-
In many regulated domains, traceability is established across diverse artifacts such as requirements, design, code, test cases, and hazards -- either manually or with the help of supporting tools, and the resulting trace links are used to support activities such as impact analysis, compliance verification, and safety inspections. Automated tracing techniques need to leverage the semantics of underlying artifacts in order to establish more accurate trace links and to provide explanations of links that have been created in either a manual or automated fashion. To support this, we propose an automated technique which leverages source code, project artifacts and an external domain corpus to generate a domain-specific concept model. We then use the generated concept model to improve traceability results and to provide explanations of the results. Our approach overcomes existing problems with deep-learning traceability algorithms, as it does not require a training set of existing trace links. Finally, as an initial proof-of-concept, we apply our semantically-guided approach to the Dronology project, and show that it improves over other tracing techniques that do not use a concept model.more » « less
-
Software traceability establishes associations between diverse software artifacts such as requirements, design, code, and test cases. Due to the non-trivial costs of manually creating and maintaining links, many researchers have proposed automated approaches based on information retrieval techniques. However, many globally distributed software projects produce software artifacts written in two or more languages. The use of intermingled languages reduces the efficacy of automated tracing solutions. In this paper, we first analyze and discuss patterns of intermingled language use across multiple projects, and then evaluate several different tracing algorithms including the Vector Space Model (VSM), Latent Semantic Indexing (LSI), Latent Dirichlet Allocation (LDA), and various models that combine mono- and cross-lingual word embeddings with the Generative Vector Space Model (GVSM). Based on an analysis of 14 Chinese-English projects, our results show that best performance is achieved using mono-lingual word embeddings integrated into GVSM with machine translation as a preprocessing step.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.1109/ICSE43902.2021.00040
