Masterarbeit
Input Data Representation Strategies in Fullscreen Chatbot Interfaces
Completion
2025/06
Research Area
Students
Daban Salahaddin Abdullah
Advisers
Verena Traubinger M.Sc.
Dr.-Ing. Sebastian Heil
Description
Chatbots have become an important interface in the ever-changing field of digital communication. They provide a conversational method for tasks like booking, data entry, and information retrieval, serving as a bridge between users and services. While chatbots offer a more natural way of interaction, their linear representation makes it hard to get an overview over data which was provided by the user. In graphical user interfaces, the current state of input data is visually represented. A combination of both approaches offers a compelling strategy to of accurately depicting input information from users in conversational interfaces, which provides them with a clear overview.
This thesis seeks to research on effective data representation within full screen chatbot interfaces, which are used for processes like purchasing or booking hotels, flights, etc, or scheduling appointments. For this, techniques have to be researched and developed that enable adequate and real-time data representation, offering users clear and concise updates on their inputs as they interact with the chatbot. This work focuses on the design of various prototypes with different techniques with which users should be able to easily review and modify their information within the chatbot conversation. These techniques differ in placement and representation style regarding the use of textual, visual, and input interface elements.
The objective of this thesis is the creation of a solution or the combination of existing approaches to solve the problem of representing and summarizing input data in fullscreen chatbots. This comprises the analysis of the state of the art of current chatbot technologies, data representation both in graphical and conversational interfaces, and other relevant literature. From this analysis, a prototypical implementation should demonstrate the feasibility of different representation strategies of the input data for a suitable use case. A suitable evaluation has to be conducted based on the effect of these different representations on the user experience and their usability, as well as its compliance with requirements which were extracted through the literature research.
