Error messages are classifiers: a process to design and evaluate error messages
| Autor: | John Wrenn, Shriram Krishnamurthi |
|---|---|
| Rok vydání: | 2017 |
| Předmět: |
Computer science
business.industry Process (engineering) 05 social sciences Perspective (graphical) 050301 education 02 engineering and technology Machine learning computer.software_genre 020204 information systems 0202 electrical engineering electronic engineering information engineering Artificial intelligence Precision and recall business 0503 education computer |
| Zdroj: | Onward! |
| DOI: | 10.1145/3133850.3133862 |
| Popis: | This paper presents a lightweight process to guide error report authoring. We take the perspective that error reports are really classifiers of program information. They should therefore be subjected to the same measures as other classifiers (e.g., precision and recall). We formalize this perspective as a process for assessing error reports, describe our application of this process to an actual programming language, and present a preliminary study on the utility of the resulting error reports. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|