English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Large-scale language models have made great progress in the field of softwareengineering in recent years. They can be used for many code-related tasks suchas code clone detection, code-to-code search, and method name prediction.However, these large-scale language models based on each code token haveseveral drawbacks: They are usually large in scale, heavily dependent onlabels, and require a lot of computing power and time to fine-tune newdatasets.Furthermore, code embedding should be performed on the entire codesnippet rather than encoding each code token. The main reason for this is thatencoding each code token would cause model parameter inflation, resulting in alot of parameters storing information that we are not very concerned about. Inthis paper, we propose a novel framework, called TransformCode, that learnsabout code embeddings in a contrastive learning manner. The framework uses theTransformer encoder as an integral part of the model. We also introduce a noveldata augmentation technique called abstract syntax tree transformation: Thistechnique applies syntactic and semantic transformations to the original codesnippets to generate more diverse and robust anchor samples. Our proposedframework is both flexible and adaptable: It can be easily extended to otherdownstream tasks that require code representation such as code clone detectionand classification. The framework is also very efficient and scalable: It doesnot require a large model or a large amount of training data, and can supportany programming language.Finally, our framework is not limited to unsupervisedlearning, but can also be applied to some supervised learning tasks byincorporating task-specific labels or objectives. To explore the effectivenessof our framework, we conducted extensive experiments on different softwareengineering tasks using different programming languages and multiple datasets.

TransformCode: A Contrastive Learning Framework for Code Embedding via Subtree transformation