Automatic Tune Family Identification by Musical Sequence Alignment

نویسندگان

  • Patrick E. Savage
  • Quentin D. Atkinson
چکیده

Musics, like languages and genes, evolve through a process of transmission, variation, and selection. Evolution of musical tune families has been studied qualitatively for over a century, but quantitative analysis has been hampered by an inability to objectively distinguish between musical similarities that are due to chance and those that are due to descent from a common ancestor. Here we propose an automated method to identify tune families by adapting genetic sequence alignment algorithms designed for automatic identification and alignment of protein families. We tested the effectiveness of our method against a high-quality ground-truth dataset of 26 folk tunes from four diverse tune families (two English, two Japanese) that had previously been identified and aligned manually by expert musicologists. We tested different combinations of parameters related to sequence alignment and to modeling of pitch, rhythm, and text to find the combination that best matched the ground-truth classifications. The best-performing automated model correctly grouped 100% (26/26) of the tunes in terms of overall similarity to other tunes, identifying 85% (22/26) of these tunes as forming distinct tune families. The success of our approach on a diverse, cross-cultural ground-truth dataset suggests promise for future automated reconstruction of musical evolution on a wide scale.

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Cs 221 Final Project: Automatic Score Following

This project aims to develop a mechanism for automatic page turning, the task of following a human performance of a musical piece, as compared to a reference musical score. At each timestep during a performance, our program outputs an estimate of the current location in the score, measured in beats from the start of the piece. The ultimate goal is to use this inference module to design an inter...

متن کامل

Social Cognition and Melodic Persistence: Where Metadata and Content Diverge

The automatic retrieval of members of a tune family from a database of melodies is potentially complicated by well documented divergences between textual metadata and musical content. We examine recently reported cases of such divergences in search of musical features which persist even when titles change or the melodies themselves vary. We find that apart from meter and mode, the rate of prese...

متن کامل

TIGRFAMs: a protein family resource for the functional identification of proteins

TIGRFAMs is a collection of protein families featuring curated multiple sequence alignments, hidden Markov models and associated information designed to support the automated functional identification of proteins by sequence homology. We introduce the term 'equivalog' to describe members of a set of homologous proteins that are conserved with respect to function since their last common ancestor...

متن کامل

BLAST for Audio Sequences Alignment: A Fast Scalable Cover Identification Tool

Searching for similarities in largemusical databases is common for applications such as cover song identification. These methods typically use dynamic programming to align the shared musical motifs between subparts of two recordings. Such music local alignment methods are slow, as are the bioinformatics algorithms they are closely related to. We have adapted the ideas of the Basic Local Alignme...

متن کامل

Bioinformatic sequence identification from sequence family databases

We have developed a tool in order to identify sequences in relation to a sequence family database. This tool combines several algorithms: BLAST, multiple sequence alignment and phylogenetic tree building. After identification of the most similar gene family to the query sequence, this query sequence is added to the whole family alignment and the phylogenetic tree of the family is rebuilt includ...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2015