Song Lyrics Summarization Inspired by Audio Thumbnailing

Proceedings of Recent Advances in Natural Language Processing, pages 328–337, Varna, Bulgaria, Sep 2–4, 2019. https://doi.org/10.26615/978-954-452-056-4_038 328 Song Lyrics Summarization Inspired by Audio Thumbnailing Michael Fell, Elena Cabrio, Fabien Gandon, Alain Giboin Universit´e Cˆote d’Azur, CNRS, Inria, I3S, France {firstname.lastname}@inria.fr Abstract Given the peculiar structure of songs, applying generic text summarization methods to lyrics can lead to the generation of highly redundant and incoherent text. In this paper, we pro- pose to enhance state-of-the-art text summa- rization approaches with a method inspired by audio thumbnailing. Instead of searching for the thumbnail clues in the audio of the song, we identify equivalent clues in the lyrics. We then show how these summaries that take into account the audio nature of the lyrics outper- form the generic methods according to both an automatic evaluation and human judgments. 1 Introduction Automatic text summarization is the task of pro- ducing a concise and fluent summary while pre- serving key information content and overall mean- ing of a text (Allahyari et al., 2017). Numerous ap- proaches have been developed to address this task and applied widely in various domains including news articles (Cheng and Lapata, 2016), scientific papers (Mei and Zhai, 2008), web content as blogs (Hu et al., 2007), customer reviews (Pecar, 2018) and social media messages (He and Duan, 2018). Just as we may need to summarize a story, we may also need to summarize song lyrics, for in- stance to produce adequate snippets for a search engine dedicated to an online song collection or for music digital libraries. From a linguistic point of view however, lyrics are a very peculiar genre of document and generic summarization methods may not be appropriate when the input for summa- rization comes from a specific domain or type of genre as songs are (Nenkova et al., 2011). Com- pared to news documents, for instance, lyrics have a very different structure. Given the repeating forms, peculiar structure (e.g. the segmentation into verse, chorus, etc.) and other unique charac- teristics of song lyrics, we need the summariza- tion algorithms to take advantage of these addi- tional elements to more accurately identify rele- vant information in song lyrics. But just as such characteristics enable the exploration of new ap- proaches, other characteristics make the applica- tion of summarization algorithms very challeng- ing, as the presence of repeated lines, the discourse structure that strongly depends on the interrelation of music and words in the melody composition, the heterogeneity of musical genres each featuring peculiar styles and wording (Brackett, 1995), and simply the fact that not all songs tell a story. In this direction, this paper focuses on the fol- lowing research questions: What is the impact of the context in summarizing song lyrics?. This question is broken down into two sub questions: 1) How do generic text summarization methods per- form over lyrics? and 2) Can such peculiar con- text be leveraged to identify relevant sentences to improve song text summarization? To answer our research questions, we experiment with generic …

Content truncated to 500 words