Download as pdf or txt
Download as pdf or txt
You are on page 1of 16

Abstract

This system will be developed to detect music plagiarism based on melodic similarity.
Melody is obtained using the harmonic structure model, and similarity between two
melodies is calculated using the edit distance. The proposed system extracts melody from
the input query and finds melodies in the Database that are close to the query melody as
potential melodies which the query has plagiarized. The proposed system is implemented
with a graphical user interface (GUI).

Keywords: Music Plagiarism, Melody Extraction, Melody Similarity.


Introduction
Music often characterized by its melody, harmony, rhythm and timbre. Amongst these
characteristics of music, melody is considered to be the most creative, specific and
expressive element that characterizes music. Though there are many kinds of plagiarism in
music, such as sample plagiarism, melody plagiarism, rhythm plagiarism and so on, melody
plagiarism is prominent in the accusation of plagiarism. Recognizing the similarities
between melodic fragments is the basic ability of a music plagiarism detection system.
Background Information
In recent years, music plagiarism has become a serious issue in the music industry.
Nowadays, with the prevalence of social media and music creation tools, musical pieces are
spreading much quickly, and music creation is getting much easier. The increasing number
of musical pieces have made the problem of music plagiarism prominent. It is easy for
common users to reach musical content anywhere and anytime on the internet while there
also exist more opportunities for unintentional and intentional plagiarism.
Problem Statement
Music plagiarism, which is the use of another work and trying to pass it off as one’s original
work, has always been a controversial topic making headlines now and then. When music
plagiarism claims arise, independent music experts will analyze the similarities between
two songs and judge them relying on their subjective opinion. This procedure is time
consuming, expensive and inefficient or error prone. Thus, a tool that can automatically
detect the similarities between songs will assist the music expert in evaluating plagiarism
quickly and effectively.
Main Objective
To develop a Music Plagiarism Detection System based on Melodic similarity to detect the
presence of similarities between a query music and music available in the Database and
decide if there is plagiarism or not.
Specific Objectives
• To develop a Melody Extraction Module to extract the melody of the query music,
• To develop a Melody-to-MIDI Module to convert the melody pitch sequence into a note
sequence which is appropriate for MIDI format,
• To develop a Database to store the music files,
• To develop a Similarity Calculation Module to calculate the similarity between the note
sequence of the input polyphonic music and those of music in the database,
• To develop a Common Subsequence Search Module to detect the similar section of the
music in the database.
Significance of the Project
• Reduce/Hinder music plagiarism cases.
• Avoidance of revenue loss due to plagiarism and pirate copies.
• To help composers avoid unnecessary plagiarism disputes
• Stimulate innovation and creativity in music industry.
Existing system
• Currently in my case study, there is no such tool used to detect plagiarism in music or
even other art works. So, the analysis of details involved in assigning copyrights to art
works especially music, is often done manually.
• Further, there are already made music plagiarism detection systems around the world,
each one with its own kind of technique used, limitations, strengths and weaknesses. One
of the technique used in existing systems is Bipartite graph matching.
• In my project, I've decided to use the melody similarity technique from the reason that,
most of our music are not notated, it is quite difficult to use some techniques without a
music score as compared to melody similarity technique. Rather, It is quite easy and
feasible to extract a melody from available polyphonic music in our country (The Case
Study).
Existing System Block Diagram

SIMPLE ANALYSIS
Song Name, genre and
Artist COSOTA some other small COPYRIGHT
details.

Independent Music
Yes PLAGIARISM
Law Experts Analysis.
CASE ARISE
Is there Plagiarism?

No

Copyright Correctly
Granted (No case)
Proposed System
The proposed system is composed of four modules as illustrated in the Figure below:
• Melody Extraction Module,
• Melody-to-MIDI Module,
• Similarity Calculation Modules and
• Common Subsequence Search Modules.
The system receives as input a polyphonic music and outputs information of plagiarized
music (music title, time, etc.).
• In the Melody Extraction Modules, melody of the query music is extracted.
• In the Melody-to-MIDI Conversion Modules, the melody pitch sequence is converted into
a note sequence which is appropriate for MIDI format.
• In Similarity Calculation Modules, the similarity between the note sequence of the input
polyphonic music and those of music in the database is calculated.
• In the Common Subsequence Search Modules, the similar section of the music in the
database is detected.
• Thus, the overall system can detect a plagiarized music and its similar section.
Methodology
Introduction.
Although the motivation for this project is to bring value to businesses related to plagiarism
in music, it is not directly linked to any business. So, a modified KDD approach is used
here. Data mining-related research/project is usually done in KDD or CRISP-DM
methodologies, but in this scenario, KDD fits best because the deployment of models in the
business layer is not applicable here, and it is not directly linked to any business.
Plagiarism in Music Methodology Approach

I will use Agile Methodology which involves breaking down a complex project into small
and manageable chunks called Sprints, which in general they build a framework known as
Scrum.
My preference to use this methodology is based on
it’s values and principles as listed below;

• Responding to change over following a plan

• Working software over comprehensive


documentation.
Sprints block diagram

SPRINT 1 SPRINT 2 SPRINT 3 SPRINT 4


● Data collection ● Database ● Model Creation ● GUI design and
● Data Pre-processing Creation ● Model Evaluation deployment
● Testing

TIME BOXING – 2 Weeks

Agile Methodology block diagram

REQUIREMENTS DESIGN DEVELOP

REVIEW DEPLOY TEST


Data Collection
Two methods will be involved in collecting data as follows;

• Interview (Semi-structured Interview)


This method encourages two-way communications.
It provides an opportunity to interviewers so that they
can learn answers to questions and the reasons behind
the answers. Further, it provides qualitative data to
compare to previous and future data.
Respondents are COSOTA officers, Music Experts and
Artists.
Cont…

• Questionnaires
These are popular research methods because they offer a
fast, efficient and inexpensive means of gathering large
amount of information from sizeable sample volumes.
These tools are particularly effective for measuring
subject behavior, preferences, intentions, attitudes and
opinions.

You might also like