A Benchmark Study on Sentiment Analysis for Software Engineering Research
This work addresses the need for reliable sentiment analysis in software engineering research, but it is incremental as it benchmarks existing tools without introducing new methods.
The paper benchmarked three sentiment analysis tools customized for software engineering to assess their performance and reliability, identifying open challenges from misclassified texts.
A recent research trend has emerged to identify developers' emotions, by applying sentiment analysis to the content of communication traces left in collaborative development environments. Trying to overcome the limitations posed by using off-the-shelf sentiment analysis tools, researchers recently started to develop their own tools for the software engineering domain. In this paper, we report a benchmark study to assess the performance and reliability of three sentiment analysis tools specifically customized for software engineering. Furthermore, we offer a reflection on the open challenges, as they emerge from a qualitative analysis of misclassified texts.