Text2PyCode: Machine Translation of Natural Language Intent to Python Source Code

Bonthu, Sridevi; Sree, S. Rama; Prasad, M. H. M. Krishna

doi:10.1007/978-3-030-84060-0_4

Cited by 4 publications

(3 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These pre-trained vectors are also added while building vocabulary for the model. The final model is shown in figure 3 taken from [11]. This model has a BLEU score of 32.40 and a ROUGE score of 85.1 which Is at par with results of natural language processing.…”

Section: Text2pycode: Machine Translationof Naturalmentioning

confidence: 98%

“…First being the baseline model which used the implementation of the Sockeye toolkit [10] containing 3 sub layers and 8 heads for multi-head attention mechanism. Next being the initial model where the Spacy tokenizer was used, here the encoder and decoder layer of the model were also fed with the word embedding created by following the Word2Vec algorithm [11]. In the final model the word embeddings of the source code was generated using the GenSim library on the CoNala python corpus.…”

Section: Text2pycode: Machine Translationof Naturalmentioning

confidence: 99%

See 1 more Smart Citation

Translation of Natural Language to Code: A Survey

Moochhala,

2021

IJEAST

View full text Add to dashboard Cite

- Learning to code in a new language takes a lot of time and effort for new as well as seasoned developers. There is a certain language that machines understand and we humans have to learn that language in order for us to communicate with the machine and make it do the things that we desire. To reduce this human-machine barrier of communication researchers have come up with a machine learning solution to translate natural language into code that allows developers to spend more time on logic and architecture rather than the actual instructions given to the machine. This paper surveys the approaches developed by various research scholars to translate natural language into machine understandable code. We analyze these approaches and state their boons and banes so that there is scope for improvement in this domain. This paper aims to give an idea about the various approaches in this domain to budding researchers wanting to contribute in this field.

show abstract

Section: Text2pycode: Machine Translationof Naturalmentioning

confidence: 98%

Section: Text2pycode: Machine Translationof Naturalmentioning

confidence: 99%

Translation of Natural Language to Code: A Survey

Moochhala,

2021

IJEAST

View full text Add to dashboard Cite

show abstract

“…Transformer Models. Bonthu et al (2021) use a standard transformer architecture (Vaswani et al, 2017) to generate Python source code from natural language intents and report BLEU score of 32.4 and Rouge-L of 82.1 on a custom curated dataset. Liang et al (2021) set a baseline for a turducken-style code generation task where SQL is embedded with Python.…”

Section: Lstm-based Seq2seq With Abstract Syntaxmentioning

confidence: 99%

A Survey on Artificial Intelligence for Source Code: A Dialogue Systems Perspective

Al-Hossami¹,

Shaikh²

2022

Preprint

View full text Add to dashboard Cite

In this survey paper, we overview major deep learning methods used in Natural Language Processing (NLP) and source code over the last 35 years. Next, we present a survey of the applications of Artificial Intelligence (AI) for source code, also known as Code Intelligence (CI) and Programming Language Processing (PLP). We survey over 287 publications and present a software-engineering centered taxonomy for CI placing each of the works into one category describing how it best assists the software development cycle. Then, we overview the field of conversational assistants and their applications in software engineering and education. Lastly, we highlight research opportunities at the intersection of AI for code and conversational assistants and provide future directions for researching conversational assistants with CI capabilities.

show abstract