DopeLearning: A computational approach to rap lyrics generation

Tutkimustuotos: Artikkeli kirjassa/konferenssijulkaisussaConference contributionScientificvertaisarvioitu

14 Sitaatiot (Scopus)

Abstrakti

Writing rap lyrics requires both creativity to construct a meaningful, interesting story and lyrical skills to produce complex rhyme patterns, which form the cornerstone of good flow. We present a rap lyrics generation method that captures both of these aspects. First, we develop a prediction model to identify the next line of existing lyrics from a set of candidate next lines. This model is based on two machine-learning techniques: the Rank SVM algorithm and a deep neural network model with a novel structure. Results show that the prediction model can identify the true next line among 299 randomly selected lines with an accuracy of 17%, i.e., over 50 times more likely than by random. Second, we employ the prediction model to combine lines from existing songs, producing lyrics with rhyme and a meaning. An evaluation of the produced lyrics shows that in terms of quantitative rhyme density, the method outperforms the best human rappers by 21%. The rap lyrics generator has been deployed as an online tool called DeepBeat, and the performance of the tool has been assessed by analyzing its usage logs. This analysis shows that machine-learned rankings correlate with user preferences.

AlkuperäiskieliEnglanti
OtsikkoKDD 2016 - Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining
KustantajaACM
Sivut195-204
Sivumäärä10
Vuosikerta13-17-August-2016
ISBN (elektroninen)9781450342322
DOI - pysyväislinkit
TilaJulkaistu - 13 elokuuta 2016
OKM-julkaisutyyppiA4 Artikkeli konferenssijulkaisuussa
TapahtumaACM SIGKDD International Conference on Knowledge Discovery and Data Mining - San Francisco, Yhdysvallat
Kesto: 13 elokuuta 201617 elokuuta 2016
Konferenssinumero: 22

Conference

ConferenceACM SIGKDD International Conference on Knowledge Discovery and Data Mining
LyhennettäKDD
MaaYhdysvallat
KaupunkiSan Francisco
Ajanjakso13/08/201617/08/2016

Sormenjälki Sukella tutkimusaiheisiin 'DopeLearning: A computational approach to rap lyrics generation'. Ne muodostavat yhdessä ainutlaatuisen sormenjäljen.

  • Siteeraa tätä

    Malmi, E., Takala, P., Toivonen, H., Raiko, T., & Gionis, A. (2016). DopeLearning: A computational approach to rap lyrics generation. teoksessa KDD 2016 - Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (Vuosikerta 13-17-August-2016, Sivut 195-204). ACM. https://doi.org/10.1145/2939672.2939679