Detection of Similar Melodies by Repurposing Algorithms for Sequence Alignment and String Searching

Project by Polygence alum Vihaan

Detection of Similar Melodies by Repurposing Algorithms for Sequence Alignment and String Searching

Project's result

Research Paper, Polygence Symposium Presentation

They started it from zero. Are you ready to level up with us?

Summary

Music plagiarism is an important concern for the music industry. Current methods of using experts to detect plagiarism are subjective and error-prone. This paper compares the performance of both string-searching algorithms and algorithms traditionally used in bioinformatics, and in particular, Knuth-Morris-Pratt (KMP) and Smith-Waterman, for the detection of melodic plagiarism. The input MIDI files are converted into an array after data processing and used as the basis for comparison. Across most thresholds, melodic plagiarism detection using KMP exhibits greater recall than, similar precision to, and faster runtimes than Smith-Waterman. We conclude that exact string searching algorithms like KMP can be more effective than local sequence alignment methods like Smith-Waterman.

Peter

Peter

Polygence mentor

PhD Doctor of Philosophy

Subjects

Computer Science, Quantitative

Expertise

machine learning for healthcare, mobile and web development for healthcare

Vihaan

Vihaan

Student

Graduation Year

2026

Project review

“All the aspects met my expectations”

About my mentor

“He gave me resources and the help I needed on the project.”