A mismatch detection system includes: a statement unit extracting portion that extracts a set of statement units by dividing a given document, which is written in a natural language, into pieces; a statement constructing portion that constructs each statement as a combination of a context and specifics by sorting each of the statement units into the context, which indicate additional information of statements, and the specifics, which indicate information of the statements; and a data generating portion that generates a data set obtained by merging a set of predetermined check specifics and a set of the statements generated by the statement constructing portion. A clustering portion converts two most similar pieces of data in the generated data set into one new piece of data which is generated by linking the most similar two pieces of data, repeats the conversion to generate a new data set, and extracts, from the generated new data set, only pieces of data that contain the statements generated by the statement constructing portion, to thereby generate a clustering result set. The mismatch detection system further includes a detection portion that generates a check item for each combination of a predetermined check subject and check specifics, and detects a mismatch of the statements based on a degree of similarity between the generated check item and a clustering result.