The discovery of microRNA (miRNA)-miRNA crosstalk has greatly improved our understanding of complex gene regulatory networks in normal and disease-specific physiological conditions. Numerous approaches have been proposed for modeling miRNA-miRNA networks based on genomic sequences, miRNA-mRNA regulation, functional information and phenomics alone, or by integrating heterogeneous data. In addition, it is expected that miRNA-miRNA crosstalk can be reprogrammed in different tissues or specific diseases. Thus, transcriptome data have also been integrated to construct context-specific miRNA-miRNA networks. In this review, we summarize the state-of-the-art miRNA-miRNA network modeling methods, which range from genomics to phenomics, where we focus on the need to integrate heterogeneous types of omics data. Finally, we suggest future directions for studies of crosstalk of noncoding RNAs. This comprehensive summarization and discussion elucidated in this work provide constructive insights into miRNA-miRNA crosstalk.