GazePrinter: Visualizing Expert Gaze to Guide Novices in a New Codebase DOI: XXXXXXX.XXXXXXX Conference: Make sure to enter the correct conference title from your rights confirmation email; June 03–05, 2018; Woodstock, NY CCS: Human-centered computing Laboratory experiments CCS: Software and its engineering Software maintenance tools CCS: Software and its engineering Software libraries and repositories CCS: Human-centered computing Human computer interaction (HCI)
Peng Kuang, Emma Söderberg, April Yi Wang, Martin Höst · 2026
Read GazePrinter: Visualizing Expert Gaze to Guide Novices in a New Codebase DOI: XXXXXXX.XXXXXXX Conference: Make sure to enter the correct conference title from your rights confirmation email; June 03–05, 2018; Woodstock, NY CCS: Human-centered computing Laboratory experiments CCS: Software and its engineering Software maintenance tools CCS: Software and its engineering Software libraries and repositories CCS: Human-centered computing Human computer interaction (HCI) by Peng Kuang, Emma Söderberg, April Yi Wang, Martin Höst (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-19855.txt.
Introduction
Kuang and co-authors test whether showing expert eye gaze helps novices learn a new codebase. Reading code is hard for beginners, yet few tools use gaze to guide them. Their GazePrinter tool turns expert gaze patterns into visual cues inside the editor. They ran a mixed study with 40 novices, using a survey, a controlled task, and interviews. Expert gaze cues changed the path novices took through the code. The authors discuss what this means for gaze-based help in developer tools.
Catalog record
| CREATOR | Peng Kuang, Emma Söderberg, April Yi Wang, Martin Höst |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 122452 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.19855
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
GazePrinter: Visualizing Expert Gaze to Guide Novices in a New Codebase DOI: XXXXXXX.XXXXXXX Conference: Make sure to enter the correct conference title from your rights confirmation email; June 03–05, 2018; Woodstock, NY CCS: Human-centered computing Laboratory experiments CCS: Software and its engineering Software maintenance tools CCS: Software and its engineering Software libraries and repositories CCS: Human-centered computing Human computer interaction (HCI) Peng Kuang, Emma Söderberg, April Yi Wang, Martin Höst Abstract . Program comprehension is an essential activity in software engineering. Not only does it often challenge professionals, but it can also hinder novices from advancing their programming skills. Gaze, an emerging modality in developer tools, has so far primarily been utilized to improve our understanding of programmers’ visual attention and as a means to reason about programmers’ cognitive processes. There has been limited exploration of integrating gaze-based assistance into development environments to support programmers, despite the tight links between attention and gaze. We also know that joint attention is important in collaboration, further suggesting that there is value in exploring collective gaze. In this paper, we investigate the effect of visualizing gaze patterns gathered from experts to novice programmers to assist them with program comprehension in a new codebase. To this end, we present GazePrinter, designed to provide gaze-orienting visual cues informed by experts to aid novices with program comprehension. We present the results of a mixed-methods study conducted with 40 novices to study the effects of using GazePrinter for program comprehension tasks. The study included a survey, a controlled experiment, and interviews. We found that visualization of expert gaze can have a significant effect on novice programmers’ behavior in terms of which path they take through the code base; with GazePrinter, novices took a path closer to the path taken by experts. We also found indications of reduced time and cognitive load among novices using GazePrinter. \keepXColumns 1. Introduction Program comprehension is a crucial task recognized by both academia and industry in software engineering (SE) . Among its variants, ranging from small single-file program to programs stretching over many files, comprehension of a program stretching over a codebase may be the most challenging due to scale and complexity. Comprehending programs of this size is not only a challenge for professional developers ( Kuang et al. 2024b ; Zhao and Tsuboi 2024 ; Green et al. 2023 ) , but also non-professional programmers such as computer science students working on group projects and scientists who need to reuse peers’ code for data analysis. In order to learn, reuse, or deploy software, programmers need to understand at least part of (if not the entirety of) a codebase. Understanding a program of the scale of a codebase adds additional challenges. The code within a shared codebase is usually co-authored and maintained by quite a few programmers from different roles, teams, and locations, as well as at different points in time. This often introduces different code styles and ways of thinking, making it inherently difficult to read and understand, especially for the first time. This situation is exacerbated as the volume of code increases. Consequently, approaching a new codebase can be time-consuming ( Green et al. 2023 ) , mentally demanding ( Gopstein et al. 2017 ; Gopstein et al. 2018 ) , and overwhelming ( Bexell et al. 2024 ) . There is a need for better tool support for program comprehension, especially as the size of the code increases. Researchers and practitioners have been developing a suite of techniques to assist in program comprehension. For example, linters (e.g. CheckStyle ( Checkstyle 2025 ) ) to synchronize style, shared documentation formats (e.g. JavaDoc ( contributors 2025b ) ) to support understandability, visualization ( Cornelissen et al. 2010 ; Hawes et al. 2015 ) to support comprehension, and walkthroughs ( Taylor and Clarke 2022 ) to guide developers through a codebase. Despite the broad adoption of these tools, many of the challenges with program comprehension remain, especially for large programs, suggesting that this is still an avenue worthy of more exploration. An idea investigated in this work is to use gaze data from experienced developers to guide developers new to a codebase in their comprehension of the codebase. Program comprehension is closely related to code reading, a task that has previously been studied using gaze analysis. However, exploration of gaze-orienting developer assistance, where gaze data gathered by an eye-tracker and analyzed as an integrated part of a developer tool, is still limited. Gaze-orienting assistance has the potential to capture subjects’ attention and to inform assistance via gaze pattern analysis. For example, Saranpää et al. ( Saranpää et al. 2023 ) present the GANDER platform with a proof-of-concept code review assistant showing the relationships between variable declarations and uses based on detected gaze patterns, and Santos et al. ( Santos 2021 ) present the Javardeye code editor supporting gaze-driven code selection and scrolling. These are examples of gaze-orienting assistance, in which gaze is analyzed during the interaction between one user and the tool. Another avenue of exploration is visualization aimed at guiding the user’s gaze. Bednarik et al. ( Bednarik et al. 2018a ) present a classroom experiment with 30 students and two single-file programs from the EMIP dataset, using an intervention in which a predefined eye movement model, defined by the program comprehension theory, guides the attention of participants during comprehension tasks. Cheng et al ( Cheng et al. 2022 ) present an experiment with 39 participants and six small single-file programs, using an intervention in which attention is guided by the gaze of other users during code review tasks. In both cases, they find positive effects of the intervention. We see a possibility to expand on these design ideas towards an approach where expert gaze gathered during a program comprehension task is encoded in a design to guide the attention of novices during the same program comprehension task. That is, use an eye movement model, similar to Bednarik et al., but informed by the gaze of others, similar to Cheng et al., specifically the gaze gathered from experts motivated by the positive effects of cognitive apprenticeship ( Collins and Kapur 2015 ) . We further see a possibility of going beyond program comprehension of single files to program comprehension of codebases, a task commonly facing novices as they enter the industry. In this paper, we present a study aimed at addressing the following research questions (RQ): RQ 1 To what extent can visualization of expert gaze assist novices in program comprehension of a new codebase? We focus on comprehension of programs on the scale and complexity of a codebase and on novices, as we see support for comprehension of larger programs as an area in need of more tool support, especially for novices who have less experience. RQ 2 To what extent do visualization of expert gaze facilitate the immediate transfer of reading strategies for novices in program comprehension of a new codebase? Given that it may be challenging to gather expert gaze for a task, it is interesting to consider if there are learning effects with regard to reading strategies between similar program comprehension tasks. RQ 3 To what extent do visualization of expert gaze influence novices’ learning experience in