Yet Another Combination of IR- and Neural-based Comment Generation

Huang, Yuchao; Wei, Moshi; Wang, Song; Wang, Junjie; Wang, Qing

doi:10.48550/arxiv.2107.12938

Cited by 3 publications

(4 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…One reason is that our method enhances the AST and code sequence, which can extract the source code semantic information more fully (explained in Section 4.2 ). LeClair et al [ 30 ] mentioned that the decoder with an attention mechanism is less effective than a transformer. We also find that the BLEU-N of the transformer-based method is on average about 7% higher than code+gnn+BiLSTM.…”

Section: Results and Analysismentioning

confidence: 99%

“…The dataset contains around 2.1 million <Java code, comment> pairs [ 29 ], which are widely used in lots of SCS generation tasks [ 10 , 20 , 30 ]. We analyzed the dataset from two aspects: (1) statistical length distribution of source codes and their comments (see Figure 9 ); and (2) a count of the scale of Java code numbers with the same comment (see Figure 10 ).…”

Section: Experiments Setupmentioning

confidence: 99%

See 1 more Smart Citation

Re_Trans: Combined Retrieval and Transformer Model for Source Code Summarization

Zhang¹,

Zhou

Qiao³

et al. 2022

Entropy

View full text Add to dashboard Cite

Source code summarization (SCS) is a natural language description of source code functionality. It can help developers understand programs and maintain software efficiently. Retrieval-based methods generate SCS by reorganizing terms selected from source code or use SCS of similar code snippets. Generative methods generate SCS via attentional encoder–decoder architecture. However, a generative method can generate SCS for any code, but sometimes the accuracy is still far from expectation (due to the lack of numerous high-quality training sets). A retrieval-based method is considered to have a higher accurac, but usually fails to generate SCS for a source code in the absence of a similar candidate in the database. In order to effectively combine the advantages of retrieval-based methods and generative methods, we propose a new method: Re_Trans. For a given code, we first utilize the retrieval-based method to obtain its most similar code with regard to sematic and corresponding SCS (S_RM). Then, we input the given code and similar code into the trained discriminator. If the discriminator outputs onr, we take S_RM as the result; otherwise, we utilize the generate model, transformer, to generate the given code’ SCS. Particularly, we use AST-augmented (AbstractSyntax Tree) and code sequence-augmented information to make the source code semantic extraction more complete . Furthermore, we build a new SCS retrieval library through the public dataset. We evaluate our method on a dataset of 2.1 million Java code-comment pairs, and experimental results show improvement over the state-of-the-art (SOTA) benchmarks, which demonstrates the effectiveness and efficiency of our method.

show abstract

Section: Results and Analysismentioning

confidence: 99%

Section: Experiments Setupmentioning

confidence: 99%

Re_Trans: Combined Retrieval and Transformer Model for Source Code Summarization

Zhang¹,

Zhou

Qiao³

et al. 2022

Entropy

View full text Add to dashboard Cite

show abstract

“…The dataset contains around 2.1 million <Java function, comment> pairs [30], which is widely used in lots of SCS generation tasks [10,22,31]. We analysis the dataset from two aspects: 1) Statistical length distribution of source codes and their comments (see Figure 8).…”

Section: Dataset Analysismentioning

confidence: 99%

Re_Trans: Combined Retrieval and transformer model for Source Code Summarization

Zhou

Qiao²,

Tang³

et al. 2022

Preprint

View full text Add to dashboard Cite

Source code summarization (SCS) refers to the natural language description on what a source code is performing. It can help developers understand programs and maintain software efficiently. Retrieval methods (RM) generate SCS reorganize terms selected from source code or use SCS of similar code snippets. Generative methods (GM) generate SCS via attentional encoder-decoder architecture. However, a GM can generate SCS for any code, but sometime the accuracy is still far from expectation (due to the lack of numerous high-quality training set). A RM is considered with a higher accuracy, but usually fail to generate SCS for a source code in the absence of a similar candidate in the database. In order to effectively combine the advantages of RM and GM, we propose a new method Re_Trans. For a given code, we first utilize RM to get the most similar code w.r.t sematic and its SCS (S_RM). Then, we input the given code and similar code into the trained discriminator. If the discriminator outputs 1, we take S_RM as the result, otherwise, we utilize the GM, transformer, to generate the given code’ SCS. Particularly, we use AST-augmented and code sequence-augmented information to make the source code semantic extraction more fully. Furthermore, we build a new SCS retrieval library through the public dataset. We evaluate our method on a data set of 2.1 million Java code-comment pairs, and experimental results show improvement over the state-of-the-art (SOTA) benchmarks, which demonstrates the effectiveness and efficiency of our method.

show abstract

“…Comment synthesis is now an active research area, including many projects such as CodeNN [30], DeepCom [26], Astattgru [40], C BERT [18], Rencos [74], SecNN [42], PLBART [1], CoTexT [54], ProphetNet-X [55], NCS [2], Code2seq [7], Re 2 Com [71], and many more [19,24,25,27,28,38,39,41,49,50,66,67,69,70,72,73]. All these approaches rely on datasets of aligned code-comment pairs.…”

Section: Introductionmentioning

confidence: 99%

Multilingual training for Software Engineering

Ahmed,

Devanbu

2021

Preprint

View full text Add to dashboard Cite

Well-trained machine-learning models, which leverage large amounts of open-source software data, have now become an interesting approach to automating many software engineering tasks. Several SE tasks have all been subject to this approach, with performance gradually improving over the past several years with better models and training methods. More, and more diverse, clean, labeled data is better for training; but constructing good-quality datasets is time-consuming and challenging. Ways of augmenting the volume and diversity of clean, labeled data generally have wide applicability. For some languages (e.g., Ruby) labeled data is less abundant; in others (e.g., JavaScript) the available data maybe more focused on some application domains, and thus less diverse. As a way around such data bottlenecks, we present evidence suggesting that human-written code in different languages (which performs the same function), is rather similar, and particularly preserving of identifier naming patterns; we further present evidence suggesting that identifiers are a very important element of training data for software engineering tasks. We leverage this rather fortuitous phenomenon to find evidence that available multilingual training data (across different languages) can be used to amplify performance. We study this for 3 different tasks: code summarization, code retrieval, and function naming. We note that this dataaugmenting approach is broadly compatible with different tasks, languages, and machine-learning models.

show abstract

Yet Another Combination of IR- and Neural-based Comment Generation

Cited by 3 publications

References 0 publications

Re_Trans: Combined Retrieval and Transformer Model for Source Code Summarization

Re_Trans: Combined Retrieval and Transformer Model for Source Code Summarization

Re_Trans: Combined Retrieval and transformer model for Source Code Summarization

Multilingual training for Software Engineering

Contact Info

Product

Resources

About