In recent years, the concept of the metaverse has evolved significantly, with the aim of defining richer immersive and interactive environments that can support various types of virtual experiences and interactions among users. This evolution has given rise to several metaverse platforms that utilize blockchain technology and non-fungible tokens (NFTs) to establish ownership of metaverse elements and attach features and information to them. This article seeks to delve into the complexity and heterogeneity of the data involved in these metaverse platforms and highlight some of the dynamics and features that make them unique. Additionally, the paper introduces a metaverse analysis tool developed by the authors, which leverages machine learning techniques to collect and analyze daily data, including blockchain transactions, platform-specific metadata, and social media trends. The experimental results of our approach are presented with a use-case scenario focused on the trading of digital parcels, commonly referred to as metaverse real estate. This scenario allows us to demonstrate the effectiveness of our tool and showcase the potential of using machine learning techniques to analyze and gain insights into the metaverse ecosystem.