Сравнение исходных текстов программ путем выравнивания…
13
The comparison of program sources using
the sequence alignment of tokens
© A.V. Dubanov
Bauman Moscow State Technical University, Moscow, 105005, Russia
Borrowing detection is a very actual problem now. In this work, one of the known algo-
rithms of the biopolymer sequence alignment was modified to make it possible to com-
pare program sources and detect similar snippets in these sources. The input data of this
algorithm are the source codes treated as the sequences of symbols. The set of lexical
domains correspond to the alphabet of symbols making up these sequences. The algo-
rithm was implemented and demonstrated with some code fragments written in Scheme
language. The perspectives and restrictions of the algorithm application are also dis-
cussed.
Keywords:
borrowing of code sequence alignment, the largest common subsequence, dy-
namic programming, lexical analysis, functional programming.
REFERENCES
[1] Burrows S., Tahaghoghi S.M.M., Zobel J. Efficient plagiarism detection for
large code repositories.
Softw. Pract. Exper
, 2007, no. 37(2), 151–175.
[2] MOSS (Measure of Software Similarity). Available at:
/ (accessed on 02.10.2014).
[3] Agrawal A., Huang X. Pairwise statistical significance of local sequence align-
ment using sequence-specific and position-specific substitution matrices.
IEEE/ACM Trans. Comput. Biol. Bioinform.
, 2011, no. 8(1), рр. 194–205.
[4] Lewis J., Ossowski S., Hicks J., Errami M., Garner H.R. Text similarity: An al-
ternative way to search MEDLINE.
Bioinformatics
, 2006, no. 22 (18), pp. 2298–
2304.
[5] Durbin R., Eddy S.R., Krogh A., Mitchison G.
Biological Sequence Analysis:
Probabilistic Models of Proteins and Nucleic Acids
. Cambridge University
Press, 1998, 336 p.
Dubanov A.V.,
Ph. D., assoc. professor of the Computer Science and Technologies De-
partment of Bauman Moscow State Technical University. Specializes in the application
of the computational methods and software development for biomedical research. He is
the author of 20 publications. e-mail:
.
