Unlocking Musicology: Digital Engagement for Digital Research

“解锁音乐学：数字研究的数字参与”将采用音乐学家的方法和分析，并表明，由于它们是数字化的，我们可以更容易地在学术界之外使用它们。这可能是为了让出版商等商业公司更好地索引或展示他们的音乐和与音乐相关的资产；或者帮助文化和慈善机构收藏我们丰富的音乐遗产，使这些更容易获得。这种重用的数字基础是在我们的前身“转换音乐学”项目中奠定的，该项目在数据、方法和结果的开发过程中采用了关联数据描述。该项目将与四个不同的合作伙伴建立四个概念验证示范，作为我们方法的示例。所有人（包括世界范围内的公众）都可以在网上使用这些工具来探索和解锁获取音乐内容的新颖而令人兴奋的方式。与我们的第一个合作伙伴纽约爱乐乐团一起，我们将为他们的Leon Levy数字档案创建关联数据，然后将语义网关系与英国类似的音乐会短暂语料库分配。我们的演示器将允许用户使用自动生成的链接无缝浏览不同的集合，例如，不同乐团的同一作品的表演，或者跟随客座指挥在合奏团之间移动。来自档案的数字项目将丰富这些探索，如生产照片，或注释的指挥乐谱。我们将与国际音乐录影带组织（RILM）合作，分析其综合国际音乐著作参考书目的内容，涵盖从19世纪初至今的出版物。在这里，我们将使用网络分析来识别兴趣和术语的社区，并提供交互式可视化，以帮助出版商和他们的客户在他们的馆藏中找到材料。除了我们帮助探索未知领域的大型收藏之外，我们还将开发新的用户界面来解释已经研究得很好的音乐。与劳伦斯·德雷福斯教授合作，我们将设计一个“数字伴侣”来说明他对瓦格纳在《罗恩格林》中使用的主题和主题的重新评价。在这里，我们将采用尖端的音乐编码和关联数据技术来突出、展示或播放德雷福斯教授叙事中的主题。同步的多媒体内容将在读者浏览文章、与评分互动或播放德雷福斯教授的解说视频时自动配置。或者，用户可以沿着自己的路径驱动体验，在浏览时检索文章、视频、音频和注释。我们的最后一个演示将与互联网档案馆合作，该档案馆是Wayback Machine和现场音乐档案馆的保管人，其中包含超过17万份高质量的现场音乐会录音。虽然计算机可以用来自动分析这些音频文件并执行基本的检索任务，但现场表演的典型特征——即兴表演、观众噪音、舞台玩笑、多变的音响效果——往往会让算法感到困惑。我们将采用一种更细致的方法，使用可以将计算分析与音乐学判断相结合的工具，来搜索乐队标准的“丢失”表演，并了解艺术家如何随着时间的推移改变他们对同一首歌的表演。在项目期间，我们将与合作伙伴分享我们的进展和见解，并传授他们如何为自己使用这些方法。在项目结束时，我们将在牛津举办一场公开展示活动，其中包括讲座和我们与合作伙伴共同开发的概念验证的实际演示。

'Unlocking Musicology: Digital Engagement for Digital Research' will take methods and analyses from musicologists and show that, because they were undertaken digitally, we can more easily make use of them beyond academia. This might be so that commercial companies, like publishers, can better index or display their music and music-related holdings; or to help cultural and charitable bodies holding rich collections of our musical heritage make these more accessible. The digital foundation for this reuse was laid in our precursor Transforming Musicology project, which adopted Linked Data descriptions during development of its data, methods and results.The project will build four proof-of-concept demonstrators, with four different partners, as examples of our approach. These will be made available online for all (including the worldwide public) to explore, and unlock novel and exciting ways of accessing musical content.With our first partner, the New York Philharmonic orchestra, we will create Linked Data for their Leon Levy Digital Archives, then assign Semantic Web relationships with similar corpora of concert ephemera in the UK. Our demonstrator will allow users to browse seamlessly between collections using automatically generated links of, for example, performances of the same work by different orchestras, or following guest conductors as they move between ensembles. Digital items from the archive will enrich these explorations, such as production photographs, or annotated conducting scores.We will work with the Répertoire International de Littérature Musicale (RILM) to analyse the content in their comprehensive international bibliography of writings on music covering publications from the early 19th century to the present. Here we will use network analyses to identify communities of interests and terms, and provide interactive visualisations to assist the publisher and their customers find material in their holdings.As well as large collections where we help explore the unknown, we will also develop novel user interfaces to explain music that is very well studied. Working with Prof. Laurence Dreyfus, we will design a 'digital companion' to illustrate his reappraisal of Wagner's use of motifs and themes in Lohengrin. Here we will deploy cutting-edge music encoding and Linked Data technology to highlight, then display or play, the motifs in Prof. Dreyfus's narrative. Synchronised multimedia content will be automatically configured as the reader moves through the essay, interacts with a score, or as a narrated video by Prof. Dreyfus is played. Alternatively the user can drive the experience along their own pathway, with the essay, video, audio, and notation being retrieved as they browse.Our final demonstrator will be in partnership with the Internet Archive, custodians of the Wayback Machine and the Live Music Archive, which contains over 170,000 high quality recordings of live concerts. While computers can be used to automatically analyse these audio files and perform basic retrieval tasks, the typical characteristics of a live show - improvisation, audience noise, stage banter, variable acoustics - often confound algorithms. We will take a more nuanced approach, using tools which can combine computational analyses with musicological judgement, to search for 'lost' performances of band standards, and see how artists change their performance of the same song over time.During the project we will share our progress and insights with our partners, and pass on how they can use these methods for themselves. At the end of the project we will host a public showcasing event in Oxford with a lecture and hands-on demonstrations of the proof-of-concepts we've developed with our partners.