Maggiori informazioni sul libro
This book summarizes various audio- and web-based approaches to music information retrieval (MIR) research, offering a holistic view that encompasses cultural and listener-centric aspects. Unlike other texts focused solely on music signal processing, it includes methods that operate on features extracted from audio signals and contextual information, such as cultural context represented online and user interactions with music. Following the document-centered paradigm of information retrieval, it discusses models of music similarity that use computational features to describe entities like songs, albums, or artists, and methods for calculating their similarities. While the representations may not capture all musical dimensions, they facilitate the discovery of music with similar qualities through abstract summarizations across different modalities. This accessible introduction to music search, retrieval, and recommendation from an academic perspective is ideal for newcomers to MIR, highlighting the music domain's relevance within the broader information retrieval community. The text is structured into four parts: Part I focuses on content-based MIR, particularly feature extraction and similarity calculation; Part II explores methods utilizing the digital cultural context of music; Part III covers collaborative filtering and user-aware retrieval; and Part IV examines current and future applications of music retrieval and recomme
Acquisto del libro
Music similarity and retrieval, Peter Knees
- Lingua
- Pubblicato
- 2016
Metodi di pagamento
Ancora nessuna valutazione.