Comunicacio presentada a la 14th International Society for Music Information Retrieval Conference, celebrada a Curitiba (Brasil) els dies 4 a 8 de novembre de 2013.

/pdf/essentia-an-audio-analysis-library-for-music-information-592ngun2yu.pdf

Essentia: An Audio Analysis Library for Music Information Retrieval.

We provide a survey of the field of Music Information Retrieval (MIR), in particular paying attention to latest developments, such as semantic auto-tagging and user-centric retrieval and recommendation approaches. We first elaborate on well-established and proven methods for feature extraction and music indexing, from both the audio signal and contextual data sources about music items, such as web pages or collaborative tags. These in turn enable a wide variety of music retrieval tasks, such as semantic music search or music identification ("query by example"). Subsequently, we review current work on user analysis and modeling in the context of music recommendation and retrieval, addressing the recent trend towards user-centric and adaptive approaches and systems. A discussion follows about the important aspect of how various MIR approaches to different problems are evaluated and compared. Eventually, a discussion about the major open challenges concludes the survey.

https://repositori.upf.edu/bitstream/10230/27565/1/Urbano_ftir_mus.pdf

Music Information Retrieval: Recent Developments and Applications

The User Requirements Notation (URN), standardized by the International Telecommunication Union in 2008, is used to model and analyze requirements with goals and scenarios. This paper describes the first ten years of development of URN, and discusses ongoing efforts targeting the next ten years. We did a study inspired by the systematic literature review approach, querying five major search engines and using the existing URN Virtual Library. Based on the 281 scientific publications related to URN we collected and analyzed, we observe a shift from a more conventional use of URN for telecommunications and reactive systems to business process management and aspect-oriented modeling, with relevant extensions to the language being proposed. URN also benefits from a global and active research community, although industrial contributions are still sparse. URN is now a leading language for goal-driven and scenario-oriented modeling with a promising future for many application domains.

User Requirements Notation: The First Ten Years, The Next Ten Years (Invited Paper)

Much work is focused upon music genre recognition (MGR) from audio recordings, symbolic data, and other modalities. While reviews have been written of some of this work before, no survey has been made of the approaches to evaluating approaches to MGR. This paper compiles a bibliography of work in MGR, and analyzes three aspects of evaluation: experimental designs, datasets, and figures of merit.

/pdf/a-survey-of-evaluation-in-music-genre-recognition-29v3l1icya.pdf

A survey of evaluation in music genre recognition

This paper presents ndings about mood representations.We aim to analyze how do people tag music by mood, tocreate representations based on this data and to study theagreement between experts and a large community. Forthispurpose,wecreateasemanticmoodspacefromlast.fmtags using Latent Semantic Analysis. With an unsuper-vised clustering approach, we derive from this space anideal categorical representation. We compare our commu-nitybasedsemanticspacewithexpertrepresentationsfromHevner and the clusters from the MIREX Audio MoodClassication task. Using dimensional reduction with aSelf-Organizing Map, we obtain a 2D representation thatwe compare with the dimensional model from Russell. Wepresent as well a tree diagram of the mood tags obtainedwith a hierarchical clustering approach. All these resultsshow a consistency between the community and the ex-perts as well as some limitations of current expert models.This study demonstrates a particular relevancy of the basicemotions model with four mood clusters that can be sum-marized as: happy, sad, angry and tender. This outcomecan help to create better ground truth and to provide morerealistic mood classication algorithms. Furthermore, thismethod can be applied to other types of representations tobuild better computational models.1. INTRODUCTIONMusicclassicationbymood

/pdf/music-mood-representations-from-social-tags-2dfmyzdvny.pdf

Music mood representations from social tags

This paper presents some findings around musical genres. The main goal is to analyse whether there is any agreement between a group of experts and a community, when defining a set of genres and their relationships. For this purpose, three different experiments are conducted using two datasets: the MP3.com expert taxonomy, and last.fm tags at artist level. The experimental results show a clear agreement for some components of the taxonomy (Blues, HipHop), whilst in other cases (e.g. Rock) there is no correlations. Interestingly enough, the same results are found in the MIREX2007 results for audio genre classification task. Therefore, a multi–faceted approach for musical genre using expert based classifications, dynamic associations derived from the wisdom of crowds, and content–based analysis can improve genre classification, as well as other relevant MIR tasks such as music similarity or music recommendation.

/pdf/the-quest-for-musical-genres-do-the-experts-and-the-wisdom-56dmp18lsi.pdf

The Quest for Musical Genres: Do the Experts and the Wisdom of Crowds Agree?

In the context of content analysis for indexing and retrieval, a method for creating automatic music mood annotation is presented. The method is based on results from psychological studies and framed into a supervised learning approach using musical features automatically extracted from the raw audio signal. We present here some of the most relevant audio features to solve this problem. A ground truth, used for training, is created using both social network information systems (wisdom of crowds) and individual experts (wisdom of the few). At the experimental level, we evaluate our approach on a database of 1,000 songs. Tests of different classification methods, configurations and optimizations have been conducted, showing that Support Vector Machines perform best for the task at hand. Moreover, we evaluate the algorithm robustness against different audio compression schemes. This fact, often neglected, is fundamental to build a system that is usable in real conditions. In addition, the integration of a fast and scalable version of this technique with the European Project PHAROS is discussed. This real world application demonstrates the usability of this tool to annotate large-scale databases. We also report on a user evaluation in the context of the PHAROS search engine, asking people about the utility, interest and innovation of this technology in real world use cases.

/pdf/indexing-music-by-mood-design-and-integration-of-an-op2z1isebg.pdf

Indexing music by mood: design and integration of an automatic content-based annotator

A robust and efficient technique for automatic music mood annotation is presented. A song's mood is expressed by a supervised machine learning approach based on musical features extracted from the raw audio signal. A ground truth, used for training, is created using both social network information systems and individual experts. Tests of 7 different classification configurations have been performed, showing that Support Vector Machines perform best for the task at hand. Moreover, we evaluate the algorithm robustness to different audio compression schemes. This fact, often neglected, is fundamental to build a system that is usable in real conditions. In addition, the integration of a fast and scalable version of this technique with the European Project PHAROS is discussed.

/pdf/music-mood-annotator-design-and-integration-3jiwzn3jj9.pdf

Music Mood Annotator Design and Integration

A key aspect of architecture-centric development is the traceability of design documentation. In particular, architects should be always aware of the relationships between the architectural model “as documented” and its corresponding implementation model. The problem is that these two models usually diverge from each other over time, due to factors such as new requirements, refactorings, etc. Therefore, tool assistance is very important to assess the level of conformance between architectural documentation and implementation. In this paper, we present a tool approach called ArchSync that helps architects to conciliate architectural documentation expressed through Use-Case Maps with Java source code, as modifications are being made on the code. ArchSync relies on a heuristic that incrementally detects inconsistencies with respect to the architectural prescriptions, based on the analysis of system execution traces. ArchSync can also give suggestions for re-synchronization. Results of two case-studies showing the applicability of the approach are reported.

Martin Blech

Papers

The Quest for Musical Genres: Do the Experts and the Wisdom of Crowds Agree?

Indexing music by mood: design and integration of an automatic content-based annotator

Music Mood Annotator Design and Integration

Assisting the synchronization of UCM-based architectural documentation with implementation