The Impact of Multimodal Inputs on User Adoption in Study Planning Applications

Abstract

With the evolution of applications, text inputs have become the most preferred form of collecting user data due to their structured form, as text input data can easily be mapped to the label keys. Technological advancements have improved the use of other input formats like images, where with the development of Optical Character Recognition tools, data can be recognised from images. Strides are underway to incorporate audio input in applications with speech recognition and synthesis tools. However, there is limited adoption of file input formats, especially Portable Document Format (PDF) files in mobile and web applications, primarily due to the highly unstructured data in PDF files. This research paper explores the role of Large Language Models in enhancing the accuracy of data collected from PDF files through structured prompting. Consequently, it explores the role of ease of data collection in influencing adoption using study applications as a case study and proposes an extension of the Technology Acceptance Model. Finally, it proposes ease of data collection as a factor influencing user's perceived ease of use of applications.