FIS-MPT: fuzzy inference system-based melody production tools. (29th October 2019)

• Record Type: Journal Article
• Authors: Rezaei, Negar; Rahmanimanesh, Mohammad
• Is Part Of: International journal of arts and technology. Volume 11:Number 4(2019)
• Page Start: 361
• Page End: 379
• Publication Date: 2019-10-29
• Subjects: music -- melody production -- fuzzy -- MIDI -- JFuzzy -- JFugue; Technology and the arts -- Periodicals; Computer art -- Periodicals; 700.285
• Languages: English
• ISSNs: 1754-8853

Abstract

One of the main concerns of artificial intelligence is to produce melody for helping new composers. This article aims to produce melody automatically by using a fuzzy inference system-based melody production tool (FIS-MPT). FIS-MPT contains three principal modules, a module for definition of fuzzy rules, a module for choosing the type of cadence in order to end the musical sentences, and a module for determination of the sentence pattern. By using FIS-MPT, a number of infinite melodies can be produced in different classes tailored to the users' tastes by changing the fuzzy inference engine, the type of cadence, and the pattern of sentences. In this regard, a software with graphical user interface is developed to produce different melodies by using Java programming environment which contains JFuzzy and JFugue libraries. Users can determine two initial notes and a tempo number. After that, the melody is automatically produced based on a default setting by using 36 fuzzy rules and authentic cadence.