Online Appendix for Semantic Dependency Visualization on Code Review: A Comparative Analysis

This online appendix provides supplementary materials associated with the study "Semantic Dependency Visualization on Code Review: A Comparative Analysis".

The files labeled P1 through P12 contain the interview transcripts for the 12 study participants, provided in the original language of the interviews, Brazilian Portuguese, to preserve the participants’ original wording, expressions, and intended meanings. The transcripts were generated automatically using the Whisper Turbo speech recognition model and subsequently reviewed to check for potentially identifying or confidential information. Speaker diarization and segmentation were also performed to distinguish the researcher’s and participant’s speech.

Because the transcripts were produced through an automated transcription and speaker-segmentation process, they may not correspond perfectly to the original audio recordings. They may contain transcription errors, omitted words, inaccurate technical terms, punctuation issues, or incorrectly assigned speaker turns. The transcripts are therefore provided primarily for documentation, transparency, and contextual reference. The qualitative analysis reported in the article was conducted using the original interview audio recordings, rather than relying exclusively on these automatically generated transcripts.

In addition to the interview transcripts, the appendix contains the following materials:

• Participant demographic summary: an anonymized overview of the participants’ development experience, review frequency, and commonly used review tools.
• Pull request descriptions: documentation of the pull requests used during the experiment.
• Interview questions: the questions and prompts used during the post-task interviews.
• Qualitative analysis codebook: the final codebook resulting from the qualitative analysis, including the codes and categories used to organize and interpret participants’ responses.
• InView interface screenshot: a screenshot showing the complete InView interface, including its dependency report, three-way diff, and dependency graph.
• InView code: a .zip file with the complete code for the tool in the same version used in the experiments.
• README: documentation for the file contents, including a description and instructions for installation and usage.
• LICENSE: the license for the artifacts.
• Paper: the camera-ready version of the paper for the study.