APIBench: A Benchmark Dataset for Evaluating API Recommendation Approaches in Python and Java

APIBench is the benchmark dataset APIBench released in the paper "Revisiting, Benchmarking and Exploring APIRecommendation: How Far Are We?". APIBench contains two sub-dataset for evaluating the performance of query-based and code-based API recommendation approaches, namely APIBench-Q and APIBench-C. Each sub-dataset has a Java version and a Python version. APIBench-Q contains 4,309 Python queries and 6,563 Java queries collected from Stack Overflow posts generated from Aug 2008 to Feb 2021 and tutorial websites Geeks4Geeks, Java2s, and Kode Java in April 2021. APIBench-C contains 2,361 Python projects and 1,477 Java projects mined from GitHub in April 2021. Please read the README.md file for detailed information about the benchmark. The evaluation results of existing API recommendation approaches can be found in this GitHub repository.

Related Organizations

Harbin Institute of Technology
China (People's Republic of)
Chinese University of Hong Kong
China (People's Republic of)

Keywords

GitHub, Tutorial Websites, Stack Overflow, Benchmark, API Recommendation, Dataset

Impact byBIP!

	selected citations These citations are derived from selected sources. This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).	0
	popularity This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.	Average
	influence This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).	Average
	impulse This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.	Average