Field data show that complex artificial fracture networks are generated along the wellbore in the Jimusar shale after large scale hydraulic fracturing. It is necessary to establish a parameter estimation method for artificial fracture networks in shale reservoirs in order to facilitate fracturing evaluation, reservoir monitoring and oil recovery improvement. Therefore, a quantitative characterization approach is first established by using the dual-porosity equivalent method to describe the complex artificial fracture networks. Based on this, a tri-linear flow-based well-test model is proposed for fracture-network wells in the Xinjiang Jimusar shale reservoir. Then the transient pressure at the wellbore is analyzed by solving the well-test model. With the transient pressure behaviors, eight fractured horizontal wells in the Jimusar shale reservoir are interpreted to obtain the parameters of the artificial fracture network. The results show that these horizontal wells have a crushed zone with fracture networks and an affected zone with higher permeability. The permeability of the reconstructed area is 130~190 mD, the width of the reconstructed area is 80~100 m, and the volume proportion of the fracture network is about 10%~14%.The width and permeability of the active zone are 90~110 m and 4~20 mD respectively. According to the variation rules of fracturing parameters and the fracturing scaleof each stage, increasing the amount of injection of each stage can improve the range and permeability of the fracturing modification area and the effective fracturing area of the Jimusar shale reservoir. When the injection volume of each stage is larger than 1900 m3, the effect of increasing the injection volume on the permeability in the reformed area is not obvious. Considering the economic benefits, the size of each stage should not exceed 1700~1900 m3. This research can provide important theoretical reference for fracturing design for the Jimusar shale reservoir in Xinjiang.