13 May 2014

Mini-projects Selected

We are delighted to be able to announce that the four Transforming Musicology mini-projects have now been selected. They are:

  • Large-scale corpus analysis of historical electronic music using MIR tools: Informing an ontology of electronic music and cross-validating content-based methods, PI: Dr Nick Collins (Durham) with Prof Peter Manning and Dr Simone Tarsitani (both Durham)
  • In Concert: Towards a Collaborative Digital Archive of Musical Ephemera, PI: Prof Rachel Cowgill (Cardiff) with Prof Alan Dix (Lancaster/Birmingham), Dr Rupert Ridgewell (British Library), Prof Simon McVeigh (Goldsmiths) and Prof Christina Bashford (University of Illinois, Urbana Champaign)
  • Medieval Music, Big Data and the Research Blend, PI: Prof Mark Everist (Southampton) with Dr Nicholas Gibbins and Dr Gregorio Bevilacqua (both Southampton)
  • Characterising stylistic interpretations through automated analysis of ornamentation in Irish traditional music recordings, PI: Dr Münevver Köküer (Birmingham City U) with Dr Peter Jančovič (Birmingham), Mr Islah Ali-MacLachlan (Birmingham CU), Prof Cham Athwal (Birmingham CU) and Dr Daithí Kearney (Dundalk Institute of Technology)

The panel felt that these four projects represent a very interesting spread, covering both very early and current musical practices, embracing both musical and extra-musical data, and engaging both senior and early career investigators. Below we reproduce the summaries of the four projects:

Large-scale corpus analysis of historical electronic music using MIR tools: Informing an ontology of electronic music and cross-validating content-based methods

Dr Nick Collins, Prof Peter Manning, Dr Simone Tarsitani

Electronic music has developed a rich history over many decades, most intensively since the second World War, with manifestations in art music and popular music spheres, and much experimental work in-between. The strong heritage of electronic music has been an increasing target of analysts, often featuring certain key works, whether Kontakte, Concret PH or Papa Sangre.

MIR tools offer the possibility to expand this endeavour to a larger database of historical recorded works, tracking audible key trends in compositional and technological endeavour, with an empirical methodology. Ironically, despite the machine-mediated creation of electronic music, automated analysis techniques have not previously been employed to any great degree. Admittedly, although inexhaustible and objective, machine audio analysis has certain limitations compared to the golden standard of human analysis. To provide grounding for this work, we would link to an existing project which is supplying human analysis of key works, taking the opportunity to validate machine methods. At the same time, we wish to explore the potential of Linked Data approaches for the release of our findings, including providing a richer context for each work within external databases of historical and social information.

The lead investigator and co-investigator would compile a larger corpus of historical works, aiming for clear coverage of important works in electronic music history, with a balanced approach to experimental art music, and popular music works. The target would be around 2000 works, a good compromise between significant corpus size and tractable calculation (if averaging four minutes per piece, this corpus size works out as around 133 hours or five and a half days of audio). We would favour the years 1948-1998, allotting fifty years from the first musique concrète to global electronic dance music before the millennium; this would suggest 400 works per decade as a target, though there would be an inevitable bulge of tracks from the 1970s to 1990s as popular music experimentation really took off and genres proliferated. The selection of works would be informed by the principal and co- investigators' experience as musicologists of electronic music, between them publishing recent key texts surveying electronic music history, and thus ideally placed as curators for this resource. They would cross reference existing texts, both their own and those of others, for the eligibility of historical pieces. In terms of feasibility, previous work of the first investigator in particular has laid the groundwork for MIR analysis of electronic music databases of more restricted historical scope (for example, synth pop, early electronic dance music). Yet the scale of the accurately annotated corpus to be created here is novel.

Over ten months, work would proceed following the gradual formation of the corpus, from small- scale experiments of prototype workflow, to large-scale computer runs over the whole database. The meta-data for the corpus, and the output data of experiments, would be encoded to be available as rich linked data online. We aim for deeper musicological insight and theory construction with our work, but further, to provide a core resource for future electronic music scholarship.

In Concert: Towards a Collaborative Digital Archive of Musical Ephemera

Prof Rachel Cowgill, Prof Alan Dix, Dr Rupert Ridgewell, Prof Simon McVeigh, Prof Christina Bashford

This project is a collaboration between four musicologists (Cowgill, Ridgewell, McVeigh, and Bashford) concerned with the creation and use of digital archives, and a technologist (Dix) with expertise in the design of human-centred systems and innovative data interfaces.

The large vision of this project is to transform processes and tools for the development, curation, and use of digital archives. We envisage a shift from a staged process that creates substantial individual authoritative resources, which are then available for research, to one where the work of curation is interleaved with the use of resources, and the power of digital technology is used to track, manage, and visualise the fine-scale progress and provenance of individual items within an evolving interlinked family of resources. This move towards technologically enabled incremental processes parallels changes in software development and web-based information systems. However, both the original recognition of the importance of this shift and our methodology in addressing it are driven bottom-up from the needs of specific datasets and musicological questions.

The datasets on which we are focusing all concern concert ephemera, but cover different periods (from the 18th to the early 20th century) and differing geographic extents (London vs all Europe), and are at different stages in the process of curation (from OCR scans to fully authoritative resource). Together these offer a musicological opportunity to augment the individual datasets and interlink them, so that questions concerning the long-term evolution of public musical performance can be addressed that no single dataset can answer on its own. However, in addition these offer a specific example of a broader issue allowing us to interrogate the different stages in archive production and inter-connection.

This bottom-up approach continues into finer detail. We will choose a small number of very specific musicological questions – for example, 'how do the sites for musical performance in London change over time in terms of geographical distribution and audience demographic?' We will address each question, using bespoke technical and human methods, but our interest in selecting appropriate questions will include the range of technical issues they raise as well as musicological importance. This, again, will have a dual aim. On the one hand these questions will allow us to achieve real musicological results within a short project. On the other hand they will act as case studies from which we can derive broader lessons about changes in professional practice and necessary technical infrastructure and tools.

This largely bottom-up approach will be complemented by top-down analysis of a more comprehensive set of curation issues and musicology questions derived partly through in-depth analysis within the team, including a research visit by Bashford, and partly by engagement with the wider community through a seminar and open discussion. Within this seminar we seek to learn from the experiences of others involved in the creation and use of digital archives, and to use the developments within this project as a stimulus for broader reflection.

As well as contributing to knowledge in Music and Computer Science research, the project will both complement and contribute to the overall aims of the Transforming Musicology project. We share a belief in the power of digital technology, when used well and appropriately, to transform musicological practice. Our focus on digital historical archives, complements the audio-analysis strengths of the Transforming Musicology project team. Our desire to interlink these archives both internally and externally will feed into the Semantic Web infrastructure of Transforming Musicology, and allow the potential for future researchers to bridge disparate areas of musicological research.

Medieval Music, Big Data and the Research Blend

Prof Mark Everist, Dr Nicholas Gibbins, Dr Gregorio Bevilacqua

The conductus is one of the central genres of medieval music, and consists of settings of Latin rhythmic poetry set to newly-composed music in the period from 1100 to around 1400 and later. Despite the survival of around 900 poems and slightly fewer musical settings, issues around the context and function of the conductus have continued to prove elusive and subject to debate and dissent. During the AHRC-funded 'Cantum pulcriorem invenire (CPI)' project, haphazard and ad hoc exploitation of big data were used largely to clarify questions associated with establishing poetic texts, their meaning and translation. One significant by-product of this work was the emergence of intertextual relationships between the texts of the conductus and other texts (a single example documents the use of the poetry of a conductus in a contemporary chronicle – completely unknown and unique in the repertory which will be written up as such). The fundamental question this project seeks to address is how far this ad hoc exploitation of big data can be systematized by the use of the Semantic Web (SW), and how far the scattered examples of intertextuality so far discovered are replicated elsewhere in the repertory.

The research materials fall into two categories: source and target data. The source data – the texts of the conductus poems – have been digitised as part of the CPI project and exist in XML format. The target data – the big data which consists of the vast amount of digitised medieval text of all forms (chronicles, technical text, patristics, poetry) – exists de facto on the web and in a variety of searchable open-access formats. One slight complication – which merely results in a slightly multiplication of the algorithms in the search engine – is that some data have been systematically edited and placed behind a paywall: Patrologia latina is a case in point; this subset of our big data will have to be handled slightly differently to other open-access material.

The identification of texts and text fragments from the source data within the much larger target dataset will provide a ranked list of correspondences between the conductus and other texts that a researcher can then process using a blend of manual and digital tools. The process outlined above is an information retrieval task, but the nature of the proposed dataset gives rise to an additional consideration: the target dataset contains errors that arise from transcription or optical character, affecting recall (some matching texts will be missed). We propose to address this using both traditional Natural Language approaches (stemming – Schinke or similar – and approximate matching) and SW technologies (representation of texts and fragments in order to match based on the topics therein).

Characterising stylistic interpretations through automated analysis of ornamentation in Irish traditional music recordings

Dr Münevver Köküer, Dr Peter Jančovič, Mr Islah Ali-MacLachlan, Prof Cham Athwal, Dr Daithí Kearney

While 'tunes' in Irish traditional music are usually of simple and regular structure, the tradition allows for and applauds the creativity of the individual musician. The perceived skill, creativity and musicality of musicians in the Irish tradition is often related to the use of ornamentation and variation in performance. The ability to accurately represent and analyse stylistic features such as ornaments allow for the development of discourse related to several key ethnomusicological questions surrounding music making, musical heritage and cultural change.

The use of terminology related to concepts of musical style may vary but usually relate to the use of identifiable features of an individual's approach to the performance of music. The style of the individual performers may be based on their experience in the tradition; listening to or playing with other traditions, perhaps in a locality or on recordings, or the use of published sheet music. While the existence of many different styles in Irish traditional music is generally accepted, an accurate and objective analysis of musical style has not yet been developed and questions remain concerning the analysis of stylistic difference in Irish traditional music. Examples of such questions could be whether there are patterns in the use of ornamentation that are favoured by a particular performer, any variations of these patterns and if exist, can they be grouped in a meaningful way to inform the understanding of musical style in Irish traditional music and to what extent patterns of musical style involving ornamentation are imitated by other performers. These questions could be extended from individual stylistic differences to regional stylistic differences and to change at various points of time through the twentieth century. Answering these questions would inform significant debates amongst both academics and practitioners related to variances in individualistic and regional styles and acknowledgement of change over time in the tradition.

In order to reliably answer these questions, a large number of recordings including various performances by musicians should be analysed, compared and contrasted. Employing computational analysis methods would enable us to analyse that large amount of recordings. In this study, we will focus on ornamentation as it is a decisive stylistic determinant in Irish traditional music. By employing state-of-the-art music information retrieval methods and/or developing novel extensions of these methods, we will be able to find what type of ornaments were used and where in the tunes they are realised. We have already developed several methods for detection of single-note ornaments. The proposed ornament detection system will be evaluated by comparing the results with the ground truth data obtained from annotation by expert listener(s). We will then investigate whether there are usage patterns of ornaments that are favoured by a particular performer and if exist, where and the frequency with which performer applies them. We will explore developing a dictionary of patterns for a particular player and make comparison across several other notable performers' recordings to ascertain whether they have unique dictionaries. While we will focus in this mini-project on understanding of individual stylistic differences, the outcomes of this project are expected to be extended to the analysis and characterisation of regional stylistic differences and change over time, which will be a part of a larger-scale project.

Outcome of the mini-project (e.g., our music collection, meta data related to manual and automated annotation of the recordings, algorithmic output, results and findings) will be shared by using the Semantic Web and the Linked Data structure and be linked to Transforming Musicology project for reuse by partners and global public.

We're looking forward to working closely with each of these project teams over the coming months. Please visit our project blog to keep up-to-date with the progress of the mini-projects and of Transforming Musicology itself.