Most private sector software development is informed by user experience (UX) designers who work to ensure products provide meaningful and relevant experiences to users, but UX is treated as an afterthought, if considered at all, within open source scientific software. The result is that much open source scientific software with graphical user interfaces is counterintuitive and difficult to use, with predictable effects on adoption. This grant funds work by Jin Guo of McGill University and Jinghui Cheng of Polytechnique Montrйal to develop and test techniques and tools to facilitate better integration of design considerations into open source software development. An essential feature of successful UX design processes is understanding user needs and soliciting and evaluating their feedback iteratively. Facilitating such activities can be difficult, particularly on open source projects where the management team is stretched thin. Drawing on their familiarity with natural language processing and human-computer interaction, Guo and Cheng will attempt to streamline user participation in UX design through three streams of research. The first will be on tools that provide prompts, suggestions and frameworks to users to reduce vagueness and ambiguity in their design feedback, clearly demarcate problems from proposed solutions, and help ensure user comments are maximally useful to developers. The second will focus on tools that use natural language recognition to analyze, summarize, and synthesize user comments, allowing the development team to digest and prioritize feedback effectively and efficiently. The third will focus on tools that engage users in contributing pre-implementation design artifacts, like wireframes, sketches, and interface mockups.