The ball penetrometer is widely performed in-situ and laboratory tests to interpret the strength characteristics of clayey soils. The penetration of ball penetrometer is a large deformation process. Therefore, large deformation finite element (LDFE) method has to be employed. However, the accuracy of the interpreted intact undrained shear strength in the strain-softening and rate-dependent nonhomogeneous soil is hampered by the penetrometer-soil interaction and soil heterogeneity. In the present study, a sliding contact model, which allows for the simulation of strain-softening effect, rate-dependent effect, and relative sliding at the interface, is proposed to simulate the penetrometer-soil interaction. The ball penetration into the seabed from the mudline is simulated using large deformation finite element analysis with the proposed interface model. The effects of dimensionless strength parameters sum/gamma ' Dandk/gamma '\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\left[ {s_{\text{um}} /\left( {\gamma<^>{\prime } D} \right){\text{and}}\;k/\gamma<^>{\prime } } \right]$$\end{document}, degree of soil strength heterogeneity (kD/sum), and interface friction coefficient (alpha) on the bearing capacity factor (N) of the ball penetrometer in strain-softening and rate-dependent soil are quantified. Finally, a back-analysis framework is proposed to interpret the intact undrained shear strength profile and the strain-softening and rate-dependent parameters of soil based on the penetration resistance of the ball penetrometer. The accuracy and robustness of the back-analysis framework are validated by comparing its results with those of numerical analysis, laboratory test and field test.