This work is dedicated to the problem of computer- aided plagiarism detection, i.e. to the development and the use of software instruments that help to reveal plagiarism. The creation of such tools raises specific algorithmic problems that deserve attention. The results covered in this work, include: (a) Building the taxonomy of existing plagiarism detection methods according to their speed and reliability characteristics. (b) Studying and improving string matching algorithms used in plagiarism detection. Introducing "tokenizers" for the natural language texts, applying natural language parsers for plagiarism detection in order to enhance the quality of the detectors. (c) Optimizing the speed performance of string matching based plagiarism detection algorithms by applying a combined fast and reliable scoring scheme. Developing an efficient parameterized matching procedure. (d) Developing a fast string matching based plagiarism detection algorithm.