Sara-Jane Bittner, from the Interactive Machine Learning department, presented a demo paper at the international conference on User Modeling, Adaptation and Personalization (ACM UMAP 2024). UMAP is an annually held premier conference for researchers and working on systems that adapt to individual users, to groups of users, and that collect, represent, and model user information. The conference was held in Cagliari, Italy from July 1st to 4th, 2024.

The paper, titled “FormTwin: A Framework for Pen-based Data Collection” focuses on a data collection tool for practitioners that bridges the gap between traditional paper-based forms and digital forms. While paper-based tools preserve input characteristics, they need to be digitized afterwards which can be time consuming and prone to errors. This can be addressed by digital forms that allow for faster processing. However, digital forms might not be suitable for target groups with low technical literacy, like elderly people. Additionally, for free-hand drawings the input quality and accuracy can be higher when using paper-based forms. FormTwin combines the two types with providing a tool with which the form can be filled out with both methods: Either directly digitally on a mobile device with a stylus or on a paper-printed form with a digital pen that transfers the input to the digital twin in real-time. That way input characteristics can be preserved, and the practitioner can choose the method of collecting data based on the target group and task characteristics of their study. Additionally, FormTwin includes an annotation tool, to which every form can be uploaded and various data types such as text, radio buttons, checkboxes but also free-hand drawings can be added.

Sara-Jane Bittner presents IML’s work at the UMAP conference in Italy (left picture) and explains the demo to a visitor (right picture)

Sara-Jane Bittner with the poster describing “FormTwin: A Framework for Pen-based Data Collection”


Konstantin Kuznetsov, Sara-Jane Bittner, Abdulrahman Mohamed Selim, Michael Barz, and Daniel Sonntag. 2024. FormTwin: A Framework for Pen-based Data Collection. In Adjunct Proceedings of the 32nd ACM Conference on User Modeling, Adaptation and Personalization (UMAP Adjunct ’24). Association for Computing Machinery, New York, NY, USA, 132–135.