Extracting and analyzing crime patterns in big cities is a challenging spatiotemporal problem. The problem's hardness is linked to the sparse nature of the crime activity and its spread in large spatial areas. Sparseness hampers most time series comparison methods from working properly, while handling large areas tends to render the computational costs of such methods impractical. Visualizing different patterns hidden on crime time series data is another issue, mainly due to the patterns that can show up from the time series analysis. In this paper, we present a new methodology that comprises two main components designed to handle the spatial sparsity and spreading of crimes in large areas. The first component relies on a stochastic mechanism to visually analyze probableXintensive crime hotspots. Such analysis reveals important patterns that can not be observed in the typical hotspot visualization. The second component builds upon a deep-learning mechanism to embed crime time series in a Cartesian-space. From the embedding, one can identify spatial locations where the crime time series have similar behavior. The two components have been integrated into a web-based analytical tool called CriPAV, enabling a global and street-level view of crime patterns. Developed in collaboration with domain experts, CriPAV has been validated through a set of case studies with real crime data. The provided experiments reveal the effectiveness of CriPAV in identifying patterns such as locations where crimes are not intense but highly probable to occur and locations that are far apart from each other but bear similar crime patterns.