Dominik Schnitzer

Location-Aware Music Artist Recommendation

Current advances in music recommendation underline the importance of multimodal and user-centric approaches in order to transcend limits imposed by methods that solely use audio, web, or collaborative filtering data. We propose several hybrid music recommendation algorithms that combine information on the music content, the music context, and the user context, in particular integrating geospatial notions of similarity. To this end, we use a novel standardized data set of music listening activities inferred from microblogs ( MusicMicro ) and state-of-the-art techniques to extract audio features and contextual web features. The multimodal recommendation approaches are evaluated for the task of music artist recommendation. We show that traditional approaches (in particular, collaborative filtering) benefit from adding a user context component, geolocation in this case.

Hybrid retrieval approaches to geospatial music recommendation

Recent advances in music retrieval and recommendation algorithms highlight the necessity to follow multimodal approaches in order to transcend limits imposed by methods that solely use audio, web, or collaborative filtering data. In this paper, we propose hybrid music recommendation algorithms that combine information on the music content, the music context, and the user context, in particular, integrating location-aware weighting of similarities. Using state-of-the-art techniques to extract audio features and contextual web features, and a novel standardized data set of music listening activities inferred from microblogs (MusicMicro), we propose several multimodal retrieval functions. The main contributions of this paper are (i) a systematic evaluation of mixture coefficients between state-of-the-art audio features and web features, using the first standardized microblog data set of music listening events for retrieval purposes and (ii) novel geospatial music recommendation approaches using location information of microblog users, and a comprehensive evaluation thereof.

Effects of album and artist filters in audio similarity computed for very large music databases

In music information retrieval, one of the central goals is to automatically recommend music to users based on a query song or query artist. This can be done using expert knowledge (e.g., www.pandora.com), social meta-data (e.g., www.last.fm), collaborative filtering (e.g., www.amazon.com/mp3), or by extracting information directly from the audio (e.g., www.muffin.com). In audio-based music recommendation, a wellknown effect is the dominance of songs from the same artist as the query song in recommendation lists. This effect has been studied mainly in the context of genre-classification experiments. Because no ground truth with respect to music similarity usually exists, genre classification is widely used for evaluation of music similarity. Each song is labelled as belonging to a music genre using, e.g., advice of a music expert. High genre classification results indicate good similarity measures. If, in genre classification experiments, songs from the same artist are allowed in both training and test sets, this can lead to over-optimistic results since usually all songs from an artist have the same genre label. It can be argued that in such a scenario one is doing artist classification rather than genre classification. One could even speculate that the specific sound of an album (mastering and production effects) is being classified. In Pampalk, Flexer, and Widmer (2005) the use of a so-called “artist filter” that ensures that a given artist’s songs are either all in the training set, or all in the test set, is proposed. Those authors found that the use of such an artist filter can lower the

A document-centered approach to a natural language music search engine

We propose a new approach to a music search engine that can be accessed via natural language queries. As with existing approaches, we try to gather as much contextual information as possible for individual pieces in a (possibly large) music collection by means of Web retrieval. While existing approaches use this textual information to construct representations of music pieces in a vector space model, in this paper, we propose a document-centered technique to retrieve music pieces relevant to arbitrary natural language queries. This technique improves the quality of the resulting document rankings substantially. We report on the current state of the research and discuss current limitations, as well as possible directions to overcome them.

A fast audio similarity retrieval method for millions of music tracks

We present a filter-and-refine method to speed up nearest neighbor searches with the Kullback–Leibler divergence for multivariate Gaussians. This combination of features and similarity estimation is of special interest in the field of automatic music recommendation as it is widely used to compute music similarity. However, the non-vectorial features and a non-metric divergence make using it with large corpora difficult, as standard indexing algorithms can not be used. This paper proposes a method for fast nearest neighbor retrieval in large databases which relies on the above approach. In its core the method rescales the divergence and uses a modified FastMap implementation to speed up nearest-neighbor queries. Overall the method accelerates the search for similar music pieces by a factor of 10–30 and yields high recall values of 95–99% compared to a standard linear search.

A model for serendipitous music retrieval

Music retrieval systems that take into account the users taste and information or entertainment need when building the results set to a query are of vital interest for academia, industry, and the passionate music listener. Unfortunately, preliminary attempts to incorporate such aspects have been rather sparse so far. Focusing on the problem of music recommendation, we therefore present a new model that combines several factors we deem to be important for personalizing retrieval results: similarity, diversity, popularity, hotness, recentness, novelty, and serendipity. We further propose different ways to measure the corresponding aspects and, where available, point to literature for a more detailed elaboration of the corresponding measures. In addition, we propose the use of social media mining techniques to address the problem of estimating popularity and hotness in a geo-aware manner.

Limitations of interactive music recommendation based on audio content

We present a study on the limitations of an interactive music recommendation service based on automatic computation of audio similarity. Songs which are, according to the audio similarity function, similar to very many other songs and hence appear unwantedly often in recommendation lists keep a significant proportion of the audio collection from being recommended at all. This problem is studied in-depth with a series of computer experiments including analysis of alternative audio similarity functions and comparison with actual download data.

A Case for Hubness Removal in High---Dimensional Multimedia Retrieval

This work investigates the negative effects of hubness on multimedia retrieval systems. Because of a problem of measuring distances in high-dimensional spaces, hub objects are close to an exceptionally large part of the data while anti-hubs are far away from all other data points. In the case of similarity based retrieval, hub objects are retrieved over and over again while anti-hubs are nonexistent in the retrieval lists. We investigate textual, image and music data and show how re-scaling methods can avoid the problem and decisively improve the overall retrieval quality. The observations of this work suggest to make hubness analysis an integral part when building a retrieval system.

One-touch access to music on mobile devices

We present an approach that offers the user a convenient and meaningful way to access her music on a mobile device. By exploiting information on acoustic similarity and community-based music labels, a music collection is automatically structured and described to allow for easy orientation and navigation within the collection. To this end, the complete collection is arranged along a circular playlist path such that similar sounding pieces are grouped together. As a consequence, regions of musical styles emerge. Furthermore, we propose two approaches to derive informative descriptors that are displayed on the different regions, allowing an overview of the whole collection at a glance. For demonstration, we implemented our prototype interface on an Apple iPod.

Choosing ℓp norms in high-dimensional spaces based on hub analysis

The hubness phenomenon is a recently discovered aspect of the curse of dimensionality. Hub objects have a small distance to an exceptionally large number of data points while anti-hubs lie far from all other data points. A closely related problem is the concentration of distances in high-dimensional spaces. Previous work has already advocated the use of fractional ℓp norms instead of the ubiquitous Euclidean norm to avoid the negative effects of distance concentration. However, which exact fractional norm to use is a largely unsolved problem. The contribution of this work is an empirical analysis of the relation of different ℓp norms and hubness. We propose an unsupervised approach for choosing an ℓp norm which minimizes hubs while simultaneously maximizing nearest neighbor classification. Our approach is evaluated on seven high-dimensional data sets and compared to three approaches that re-scale distances to avoid hubness.