Abstract: Textual similarity calculation is one of the most important tasks of natural language processing. By summing up and analyzing the classical methods and the latest developments, the text representation and textual similarity calculation in deep learning were summarized and reviewed comprehensively. Due to the development of social networks, short textual similarity calculation, an important subtask has been derived. For each module, this paper sorted out the related technologies and theoretical foundations of textual similarity calculation, and summarized their specific applications and improvements in short texts. This paper summed up the common data sets and evaluation indexes in the field of textual similarity calculation. Some possible directions of textual similarity calculation were pointed out.