Recognizing causal elements and causal relations in text is one of the\nchallenging issues in natural language processing; specifically, in low\nresource languages such as Persian. In this research we prepare a causality\nhuman annotated corpus for the Persian language which consists of 4446\nsentences and 5128 causal relations and three labels of cause, effect and\ncausal mark -- if possibl -- are specified for each relation. We have used this\ncorpus to train a system for detecting causal elements boundaries. Also, we\npresent a causality detection benchmark for three machine learning methods and\ntwo deep learning systems based on this corpus. Performance evaluations\nindicate that our best total result is obtained through CRF classifier which\nhas F-measure of 0.76 and the best accuracy obtained through Bi-LSTM-CRF deep\nlearning method with Accuracy equal to %91.4.\n