Automatic Algorithm Programming Model Based on the Improved Morgan's Refinement Calculus

Zuo, Zhentao; Hu, Ying; Huang, Qing; Yuan, Wei; WANG, Changjing

doi:10.1051/wujns/2022275405

Cited by 2 publications

(2 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The method emphasizes problem induction prior to construction and proposes efficient strategies for the sequential statistics class of problems during the induction process, yielding efficient inductive recurrence relations. The Morgan 􀆳 s refinement calculus is improved in the construction [7] . The inductive recursive relation is used as a guide to gradually refine the program to obtain a highly reliable abstract program, which is then converted into an executable C++ program via the conversion platform.…”

Section: Introductionmentioning

confidence: 99%

Program Construction Method for Sequential Statistics Class Algorithm Based on Bidirectional Scanning Induction

ZUO,

WANG,

LIANG

et al. 2023

Wuhan Univ. J. Nat. Sci.

Self Cite

View full text Add to dashboard Cite

The program construction process is based on rigorous mathematical reasoning, which leads to a fully correct algorithmic program via step-by-step refinement of the program specifications. The existing program construction methods' refinement process is partly based on individual subjective speculation and analysis, which lacks a precise guidance method. Meanwhile, efficiency factors have usually been ignored in the construction process, and most of the constructed abstract programs cannot be run directly by machines. In order to solve these problems, a novel program construction method for the sequence statistical class algorithms based on bidirectional scan induction is proposed in this paper. The method takes into account the efficiency factor and thus improves the Morgan's refinement calculus. Furthermore, this paper validates the method's feasibility using an efficiency-sensitive sequential statistics class algorithm as a program construction example. The method proposed in this paper realizes the correctness construction process from program specifications to efficient executable programs.

show abstract

Section: Introductionmentioning

confidence: 99%

Program Construction Method for Sequential Statistics Class Algorithm Based on Bidirectional Scanning Induction

ZUO,

WANG,

LIANG

et al. 2023

Wuhan Univ. J. Nat. Sci.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Code comments play a crucial role in software development and maintenance, aiding developers to comprehend and reuse code without meticulous examination [1][2][3][4][5][6] . For example, in the process of transforming program specifications into code, code summarization can significantly enhance the readability and comprehensibility of the code [2][3][4][5] . However, manually composing code comments for extensive programming projects can be labor-intensive and time-consuming.…”

Section: Introductionmentioning

confidence: 99%

Improve Code Summarization via Prompt-Tuning CodeT5

2023

Wuhan Univ. J. Nat. Sci.

View full text Add to dashboard Cite

Code comments are crucial in software engineering, aiding in program maintenance and code reuse. The process of generating clear and descriptive code comments, outlining code functionality, is called code summarization. Existing code summarization methods are typically trained using transformer-based models. However, these trained models often possess limited parameters and lack specific training tasks, hindering their ability to capture code semantics effectively. This paper uses a high-capacity pre-trained model, CodeT5, for code summarization. CodeT5 is designed with an encoder-decoder architecture that excels in code summarization tasks. Furthermore, we adopt a novel paradigm, "pre-train, prompt, predict", to unlock the knowledge embedded within CodeT5. We devise a prompt template to convert input code into code prompts and fine-tune CodeT5 with these prompts—a process we term prompt tuning. Our effectiveness experiments demonstrate that prompt tuning CodeT5 with only 40% of the dataset can achieve comparable performance to fine-tuning CodeT5 with 100% of the dataset. This means our approach is applicable in few-shot learning scenarios. Additionally, our prompt learning method is not sensitive to the size of the tuning dataset. Our practicality experiments show that the performance of prompt-tuned CodeT5 far surpasses that of transformer-based models trained on code-comment datasets collected from Stack Overflow.

show abstract

Automatic Algorithm Programming Model Based on the Improved Morgan's Refinement Calculus

Cited by 2 publications

References 12 publications

Program Construction Method for Sequential Statistics Class Algorithm Based on Bidirectional Scanning Induction

Program Construction Method for Sequential Statistics Class Algorithm Based on Bidirectional Scanning Induction

Improve Code Summarization via Prompt-Tuning CodeT5

Contact Info

Product

Resources

About