This study was designed to examine whether people's expectations differ regarding how music lyrics affect individual behavior as a function of music genre. Because legislative attention and media publicity have been biased against certain types of popular music (i.e., heavy metal and rap), the authors expected that those genres of music would be viewed more negatively than other genres of popul...

Understanding the mood of music holds great potential for recommendation and genre identification problems. Unfortunately, hand-annotating music with mood tags is usually an expensive, time-consuming and subjective process, to such an extent that automatic mood recognition methods are required. In this paper we present a new unsupervised learning approach for mood recognition, based on the lyri...

Imperfect and internal rhymes are two important features in rap music previously ignored in the music information retrieval literature. We developed a method of scoring potential rhymes using a probabilistic model based on phoneme frequencies in rap lyrics. We used this scoring scheme to automatically identify internal and line-final rhymes in song lyrics and demonstrated the performance of thi...

We adapt the hierarchical attention network for the task of genre classification using lyrics. Utilising a large dataset of intact song lyrics we apply a recurrent neural network model which tries to learn importance of words, lines and sections in the genre classification task. This hierarchical structure attempts to replicate the structure of lyrics and enable learning of which sections, line...

Recent research in cognitive science and psycholinguistics has provided evidence that extreme behaviors such as suicide may be predicted by analyzing an individual’s writing style. This study seeks to extend research of this kind to a new and less traditional written genre: song lyrics. We applied the computational tools Coh-Metrix and Linguistic Inquiry and Word Count to a corpus of 35 songs e...

Part-of-speech (POS) taggers trained on newswire perform much worse on domains such as subtitles, lyrics, or tweets. In addition, these domains are also heterogeneous, e.g., with respect to registers and dialects. In this paper, we consider the problem of learning a POS tagger for subtitles, lyrics, and tweets associated with African-American Vernacular English (AAVE). We learn from a mixture o...

Infants are often presented with input in which there are multiple related regularities, as is the case in musical input with both melodic and lyrical structure. Adult learners often learn more easily from complex input containing multiple correlated regularities than from simplified input. Do infants also capitalize on complexity, or instead do they benefit from simplified input? In this serie...

Music and speech are two realms predominately species-specific to humans, and many human creative endeavors involve these two modalities. The pairing of music and spoken text can heighten the emotional and cognitive impact of both the complete song being much more compelling than either the lyrics or the accompaniment alone. This work describes a system that is able to automatically generate an...