“Show Me What’s Wrong!”: Combining Charts and Text to Guide Data Analysis
Beatriz Feliciano - Feedzai, Lisbon, Portugal
Rita Costa - Feedzai, Lisbon, Portugal
Jean Alves - Feedzai, Porto, Portugal
Javier Liébana - Feedzai, Madrid, Spain
Diogo Ramalho Duarte - Feedzai, Lisbon, Portugal
Pedro Bizarro - Feedzai, Lisbon, Portugal
Download camera-ready PDF
Room: Bayshore II
2024-10-14T16:00:00ZGMT-0600Change your timezone on the schedule page
2024-10-14T16:00:00Z
Fast forward
Abstract
Analyzing and finding anomalies in multi-dimensional datasets is a cumbersome but vital task across different domains. In the context of financial fraud detection, analysts must quickly identify suspicious activity among transactional data. This is an iterative process made of complex exploratory tasks such as recognizing patterns, grouping, and comparing. To mitigate the information overload inherent to these steps, we present a tool combining automated information highlights, Large Language Model generated textual insights, and visual analytics, facilitating exploration at different levels of detail. We perform a segmentation of the data per analysis area and visually represent each one, making use of automated visual cues to signal which require more attention. Upon user selection of an area, our system provides textual and graphical summaries. The text, acting as a link between the high-level and detailed views of the chosen segment, allows for a quick understanding of relevant details. A thorough exploration of the data comprising the selection can be done through graphical representations. The feedback gathered in a study performed with seven domain experts suggests our tool effectively supports and guides exploratory analysis, easing the identification of suspicious information.