Abstract. This paper proposes an ensemble method for multilabel classification. The RAndom k-labELsets (RAKEL) algorithm constructs each member of the ensemble by considering a small random subset of labels and learning a single-label classifier for the prediction of each element in the powerset of this subset. In this way, the proposed algorithm aims to take into account label correlations using single-label classifiers that are applied on subtasks with manageable number of labels and adequate number of examples per label. Experimental results on common multilabel domains involving protein, document and scene classification show that better performance can be achieved compared to popular multilabel classification approaches. 1

A large body of research in supervised learning deals with the analysis of singlelabel data, where training examples are associated with a single label λ from a set of disjoint labels L. However, training examples in several application domains are often associated with a set of labels Y ⊆ L. Such data are called multi-label.

There is a growing interest in developing and then evaluating Music Information Retrieval (MIR) systems that can provide automated access to the mood dimension of music. Mood as a music access feature, however, is not well understood in that the terms used to describe it are not standardized and their application can be highly idiosyncratic. To better understand how we might develop methods for comprehensively developing and formally evaluating useful automated mood access techniques, we explore the relationships that mood has with genre, artist and usage metadata. Statistical analyses of term interactions across three metadata collections (AllMusicGuide.com, epinions.com and Last.fm) reveal important consistencies within the genre-mood and artist-mood relationships. These consistencies lead us to recommend a cluster-based approach that overcomes specific term-related problems by creating a relatively small set of data-derived “mood spaces ” that could form the ground-truth for a proposed MIREX “Automated Mood Classification ” task.

The ever-growing amount of available music induces an increasing demand for Music Information Retrieval (MIR) applications such as music recommendation applications or automatic classification algorithms. When audio-based, a crucial part of such systems are the audio feature extraction routines. In this paper, we evaluate how well a variety of combinations of feature extraction and machine learning algorithms are suited to classify music into perceptual categories. The examined categorizations are perceived tempo, mood (happy / neutral /sad), emotion (soft / neutral / aggressive), complexity, and vocal content. The aim is to contribute to the investigation which aspects of music are not captured by the common audio descriptors; from our experiments we can conclude that most of the examined categorizations are not captured well. This indicates that more research is needed on alternative (possibly extra-musical) sources of information for useful music classification. 1.

Abstract. The increased popularity of tagging during the last few years can be mainly attributed to its embracing by most of the recently thriving user-centric content publishing and management Web 2.0 applications. However, tagging systems have some limitations that have led researchers to develop methods that assist users in the tagging process, by automatically suggesting an appropriate set of tags. We have tried to model the automated tag suggestion problem as a multilabel text classification task in order to participate in the ECML/PKDD 2008 Discovery Challenge. 1

A standardized mood classification testbed is needed for formal cross-algorithm comparison and evaluation. In this poster, we present a simplification of the problems associated with developing a ground-truth set for the evaluation of mood-based Music Information Retrieval (MIR) systems. Using a dataset derived from Last.fm tags and the USPOP audio collection, we have applied a K-means clustering method to create a simple yet meaningful cluster-based set of high-level mood categories as well as a ground-truth dataset. 1

I hereby declare that I am the sole author of this thesis. This is a true copy of the thesis, including any required final revisions, as accepted by my examiners. I understand that my thesis may be made electronically available to the public. ii The goal of this project is to apply system identification techniques to model people’s perception of emotion in music as a function of time. Emotional appraisals of six selections of classical music are measured from volunteers who continuously quantify emotion using the dimensions valence and arousal. Also, features that communicate emotion are extracted from the music as a function of time. By treating the features as inputs to a system and the emotional appraisals as outputs of that system, linear models of the emotional appraisals are created. The models are validated by predicting a listener’s emotional appraisals of a musical selection (song) unfamiliar to the system. The results of this project show that system identification provides a means to improve previous models for individual songs by allowing them to generalize emotional appraisals for a genre of music. The average

Abstract. Music is not only a set of sounds, it evokes emotions, subjectively perceived by listeners. The growing amount of audio data available on CDs and in the Internet wakes up a need for content-based searching through these files. The user may be interested in finding pieces in a specific mood. The goal of this paper is to elaborate tools for such a search. A method for the appropriate objective description (parameterization) of audio files is proposed, and experiments on a set of music pieces are described. The results are summarized in concluding chapter. 1

Abstract not found

This paper surveys the state of the art in automatic emotion recognition in music. Music is oftentimes referred to as a “language of emotion ” [1], and it is natural for us to categorize music in terms of its emotional associations. Myriad features, such as harmony, timbre, interpretation, and lyrics affect emotion, and the mood of a piece may also change over its duration. But in developing automated systems to organize music in terms of emotional content, we are faced with a problem that oftentimes lacks a welldefined answer; there may be considerable disagreement regarding the perception and interpretation of the emotions of a song or ambiguity within the piece itself. When compared to other music information retrieval tasks (e.g., genre identification), the identification of musical mood is still in its early stages, though it has received increasing attention in recent years. In this paper we explore a wide range of research in music emotion recognition, particularly focusing on methods that use contextual text information (e.g., websites, tags, and lyrics) and content-based approaches, as well as systems combining multiple feature domains. 1.