Understanding User Engagement With Multi-Representational License Comprehension Interfaces

Introduction

Software developers routinely encounter legal documents such as regulatory controls for security and safety, privacy policies, and open-source licenses during software development and maintenance. Without appropriate training, it can be a challenge to understand the complex structures in legal documents and fully comprehend legal jargon. The term “legalese” has been used to describe the technical and formal language used in a legal document that is often difficult to convey in simple English. As open source engagement and development can be found in most organizations today, failure to understand legalese by software developers can introduce compliance risks. Licenses being at the core of any open-source software exchange, incorrect or inconsistent interpretation of legalese can lead to litigation. To address these risks, recent studies suggest the need for tool support to help guide developers in understanding single and multi-license interactions. However, these studies do not examine how developers use sections or even sentences within a single license in exploring a compliance problem. Besides, new methods are now available to visualize rights and obligations that are fundamental to the legal structure of an open-source software license and its interpretation by various stakeholders (Mandal, Gandhi, Siy 2020). Consequently, this paper designs an eye-tracking study to understand how developers comprehend licenses using a multi-representational license exploration tool. Eye-tracking is often used in scientific studies to investigate the subject's cognitive process, especially in problem-solving tasks. Thus, using eye trackers in user studies can provide important insights for the design of interactive, multi-representational license comprehension tool interfaces. Our experiment setup allows the study of interaction patterns used by a developer when understanding the license text for exploring a compliance problem scenario. Prior research using survey-based methods claim that developers comprehend legal issues, but it is not clear how a tool can improve or assess the depth of their understanding. To better understand tool use, new studies are required to capture and analyze user interactions resulting from the presence of different modalities of the license text. To support such studies, we have developed a repeatable eye tracking-based methodology for multi-representational license comprehension interfaces.