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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Your computer—through which you are, in all likelihood, reading this paper—is continuously, efficiently, and reliably executing computer programs, but does it really understand them? Not in any meaningful sense. That burden falls upon human knowledge workers, who are increasingly asked to write and understand code. They would benefit greatly from intelligent tools that reveal the connections between their code, their colleagues’ code, and the subject-matter concepts to which the code implicitly refers and to which their real enthusiasm belongs. By teaching machines to comprehend code, we could create artificial agents that empower human knowledge workers or perhaps even generate useful programs of their own. One computational domain undergoing rapid growth is data science. Besides the usual problems facing the scientistturned-programmer, the data scientist must contend with a proliferation of programming languages (like Python, R, and Julia) and frameworks (too numerous to recount). Data science therefore presents an especially compelling target for machine understanding of computer code. An AI agent that simultaneously comprehends the generic concepts of computing and the specialized concepts of data science could prove enormously useful, for example to debug and visualize machine learning workflows or automatically summarize data analyses as natural text for human readers. Towards this vision, we propose and implement an AI system that forms semantic representations of computer programs in a particular subject-matter domain. We will focus on applications to data science because we, the authors, are all data scientists of various stripes. Nevertheless, we think that our methodology could be fruitfully applied to other scientific domains with a heavy computational focus, such as bioinformatics or computational linguistics.

Semantic Representation of Data Science Programs