Oil and gas shales are a class of multiscale, multiphase, hybrid inorganic-organic sedimentary rocks that consist of a generally uniform, preferentially oriented clay matrix with randomly embedded silt and sand particles as solid inclusions. A thorough understanding of the mechanical properties of shales is crucial for the exploration and production of oil and gas in the unconventional shale reservoirs, but it can be a challenging task due to their nature of compositional heterogeneity and microstructural anisotropy. In efforts to better characterize the mechanical properties of shales across different length scales and to fundamentally understand the laws of upscaling from individual constituent minerals to bulk rocks, a big data-based nanoindentation technique was proposed and validated in the first phase of this research. This technique is capable of simultaneously measuring the properties of both clay matrix and individual solid inclusions at the nanoscale and the properties of bulk rock at the macroscale, which established the theoretical foundation of this study and provided a powerful tool for the future research on characterizing the cross-scale mechanical properties of other composite materials (e.g., concretes). Sufficient data obtained from the past and existing operations have suggested that the fracturing fluids can cause the weakening of rock frame in the shale formations during the hydrofracking processes, but the mechanisms behind this softening behavior still remain elusive. Therefore, the second phase of this research investigated the effects of various fluids used in oil-field practice on the mechanical properties of shale by implementing the big data-based nanoindentation technique. Together with some chemical characterization methods, it was found that the water-based fracturing fluids usually dissolved the carbonate cementation in shale via some physical interactions and chemical reactions, which reduced the microstructural integrity of shale and hence the mechanical performances. However, the indentation results also confirmed that the shale softening can be effectively suppressed after adding a certain surfactant, indicating that the big data-based nanoindentation can serve as a useful tool to optimize the selection of fracturing fluid for a given shale formation. In the final phase of this study, the effect of different interlayer complexes (e.g., interlayer cation species and layer charges) on the elasticity of smectite, a swelling clay mineral commonly found in shales, was further investigated by using the nanoindentation testing. The findings here can help understand the elasticity alteration of clay matrix in bulk shales upon contact with fracturing fluids, which was achieved through elucidating the upscaling correlation from individual smectite crystals to assemblage of clay particles (i.e., clay matrix).