Problem Description
User-centered design techniques for creating innovative and user-oriented products and services are being adopted by a heavily increasing number of companies, from small businesses to global players. But despite the many benefits that such approaches offer, their usage also comes with challenges. The direct personal contact and exchange with users that is required for many design techniques is expensive, time-consuming, and does not scale, especially for the early phases of the design process. Speech-based conversational agents such as Amazon’s Alexa have the potential to help generating data more efficiently and therefore enabling the application of design techniques on a large scale. Currently, the usage of conversational agents for user-centered design processes is largely unexplored. A prominent example of the scalability issue is the application of interview techniques within design processes. Even though collaboration tools such as Skype allow for face-to-face longdistance interviews, the setup is challenging and the number of interviews that a team is able to conduct as part of a user-centered design process is limited. As such, this master thesis aims at exploring the feasibility of using conversational agents for the automated conduction of interviews, and its underlying requirements.
Goal of this Thesis
Following the design science research methodology, the main goal of the thesis is the identification of meta-requirements and design principles for the conduction of interviews with the help of speech-based conversational agents, and the prototypical implementation of an Alexa skill. The prototype will be evaluated following a qualitative research approach.
The student will have to work on the following questions:
- What is the status quo of conversational agents? Where have they been applied in the context of user-centered design?
- How are interviews in design processes conducted currently and which parts can be performed by conversational agents?
- How can an Alexa-based prototype look like based on the identified requirements?
- How can the quality of interview interactions with such agents be assessed and improved?
Skills Required
- Very good English skills (the thesis will be written in English)
- Experience with coding environments (preferably Javascript / Java, Alexa Skills Kit is beneficial)
- Experience in project environments
- General interest in human-computer interaction
- Experience in cross-disciplinary project teams
- Knowledge of design techniques is beneficial
- Excellent time management and self-organization skills
- Interest in product development, agile and user-centered development approaches
Partner
The master thesis is integrated into the Agile Innovation Lab 2018 project, hosted by the Institute of Product Engineering (IPEK). Over the course of 6 months, students will be integrated into a multidisciplinary team of 5-7 members (e.g. from mechanical engineering, industrial engineering & management or information engineering and management) to solve a real-world product development challenge proposed by a well-known industry partner. As part of the development cycle, the team will apply agile methods, such as SCRUM, over the course of six project phases with individual milestones. As a project team, students identify user needs, analyze potential solutions, design concepts and build and evaluate prototypes, with the help of a variety of design techniques. At the same time, the students utilize their project results and experiences as a source of data for their individual thesis topics. At the end of the project, the team will be invited to present their findings at a colloquium with the industry partner company and get the chance to present their results in front of an expert committee.
The student team receives individual coaching as well as workshops on scientific methods and product engineering from supervisors at IPEK, alongside the thesis supervision from the Institute of Information Systems and Marketing (Research Group Information Systems and Service Design, Prof. Mädche), with support from the Karlsruhe Service Research Institute (KSRI).
[Image credit: amazon.jobs]
