Last.fm is a website of music that provides a multivariate dataset with complex relationships between the data items and data tables. However, the relationships are difficult to examine if using purely Last.fm APIs and its Web services. In this paper, we develop a visualization technique, called Hyper Word Clouds, for the examination of such complex and multi-relationship dataset. Through the text-based representation, tracks, albums, artists, and other Last.fm data items are visualized as words linked in parallel and anchorbased word clouds. The users can then interactively select to filter, highlight, and compare data and relationships of interest and to discover further insights. CCS Concepts •Human-centered computing → Visualization techniques; Information visualization; Keywords Data relationship, music collection, text visualization