While predictive policing has become increasingly common in assisting with decisions in the criminal justice system, the use of these results is still controversial. Some software based on deep learning lacks accuracy (e.g., in F-1), and many decision processes are not transparent causing doubt about decision bias, such as perceived racial, age, and gender disparities. This paper addresses bias issues with post-hoc explanations to provide a trustable prediction of whether a person will receive future criminal charges given one's previous criminal records by learning temporal behavior patterns over twenty years. Bi-LSTM relieves the vanishing gradient problem, and attentional mechanisms allows learning and interpretation of feature importance. Our approach shows consistent and reliable prediction precision and recall on a real-life dataset. Our analysis of the importance of each input feature shows the critical causal impact on decision-making, suggesting that criminal histories are statistically significant factors, while identifiers, such as race, gender, and age, are not. Finally, our algorithm indicates that a suspect tends to gradually rather than suddenly increase crime severity level over time.