Abstract
Unstructured information artefacts, such as social media posts, news articles, and company reports, are vital for knowledge workers but often suffer from uncertain quality, making evaluation challenging. Metadata can serve as a valuable aid in assessing data quality, yet little research explores how users engage with it. This study investigates user interaction with metadata when evaluating large PDF documents. Through qualitative analysis, we examine user behaviours, cognitive processes, and trust dynamics in metadata-driven assessments. Our findings reveal how users explore, verify, and utilize metadata to assess quality, highlighting its alignment with dimensions such as relevance, completeness, and timeliness. We identify key factors influencing trust in metadata and its role as a cognitive aid in decision-making. Our results contribute to a deeper understanding of metadata’s impact on data quality assessment and inform the design of user-centred systems that support more effective decision-making in unstructured data environments.
Recommended Citation
Zhou, Hui; Chen, Tianwa; Demartini, Gianluca; Indulska, Marta; and Sadiq, Shazia, "“We Don’t Have Time to Read the Entire Document to Evaluate
the Quality” – The Role of Metadata in Quality Evaluation
of Unstructured Textual Data" (2025). ACIS 2025 Proceedings. 82.
https://aisel.aisnet.org/acis2025/82