DOCSLIB.ORG

Knolwedge Acquisi$On from Music Digital Libraries

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Knolwedge Acquisi$On from Music Digital Libraries Knolwedge Acquision from Music Digital Libraries Sergio Oramas, Mohamed Sordo Mo0vaon • Musicological Knowledge is hidden between the lines • Machines don’t know how to read Why Knowledge Acquisi0on? • Obtain knowledge automacally • Make complex ques0ons • Visualize the informaon • Improve navigaon • Share knowledge 4 Musical Libraries Musical Libraries digital recording, scan Musical Libraries OCR, manual transcrip0on digital recording, scan Musical Libraries Informaon Extrac0on, seman0c annotaon OCR, manual transcrip0on digital recording, scan Musical Libraries Informaon Extrac0on, seman0c annotaon Current DL OCR, manual transcrip0on digital recording, scan Musical Libraries Web search Informaon Extrac0on, seman0c annotaon Current DL OCR, manual transcrip0on digital recording, scan Musical Libraries Web search Informaon Extrac0on, seman0c annotaon Current DL OCR, manual transcrip0on digital recording, scan Seman0c Web • The Semanc Web aims at conver0ng the current web, dominated by unstructured and semi-structured documents into a web of linked data. • Achievements useful for Digital Libraries – Common framework for data representaon and interconnec0on (RDF, ontologies) – Seman0c technologies to annotate texts (En0ty Linking) – Language for complex queries (SPARQL) Wikipedia and DBpedia - Digital Encyclopedia - Knowledge Base - Unstructured - Structured - Keyword search - Query search 13 14 15 DBpedia • Dbpedia example queries – Composers born in Vienna in XVIII Century – American jazz musicians that have wri\en songs recorded by RCA Records • Dbpedia graph applicaons – En0ty Relevance – En0ty Similarity – En0ty Recommendaon 16 Google Knowledge Graph 17 Music Digital Libraries • Current navigaon – Indexes – Keyword-based search • From searchable repositories to knowledge environments – Complex queries – Informaon visualizaon – Recommendaon – Data Analy0cs – Q&A Music Digital Libraries 19 Music Digital Libraries 20 Dataset: The New Grove • Encyclopedic dic0onary • One of the largest reference works in Western music • Ar0st biographies crawled from the Grove Music Online – 16,707 biographies (1st paragrahph) – From pre-medieval to contemporary Dataset: The New Grove • What are the most relevant music schools? • What are the ar0sts most similar to Schoenberg? • Which are the most represented roles in the Grove? • Is there a migraon tendency in ar0sts? To which ci0es? • What is the best city to die for a musician? 22 Dataset: The New Grove 23 Informaon Extrac0on Anton Webern (b Vienna, 3 Dec 1883; d Mi\ersill,15 Sept 1945). Austrian composer and conductor. Webern, who was probably Schoenberg's first private pupil, and Alban Berg, who came to him a few weeks later […] Boulez and Stockhausen and other integral serialists of the Darmstadt School […] 24 Informaon Extrac0on Anton Webern (b Vienna, 3 Dec 1883; d Mi\ersill,15 Sept 1945). Austrian composer and conductor. Webern, who was probably Schoenberg's first private pupil, and Alban Berg, who came to him a few weeks later […] Boulez and Stockhausen and other integral serialists of the Darmstadt School […] 25 Informaon Extrac0on Anton Webern (b Vienna, 3 Dec 1883; d Mi\ersill,15 Sept 1945). Austrian composer and conductor. Webern, who was probably Schoenberg's first private pupil, and Alban Berg, who came to him a few weeks later […] Boulez and Stockhausen and other integral serialists of the Darmstadt School […] 26 Informaon Extrac0on Anton Webern (b Vienna, 3 Dec 1883; d Mi\ersill,15 Sept 1945). Austrian composer and conductor. Webern, who was probably Schoenberg's first private pupil, and Alban Berg, who came to him a few weeks later […] Boulez and Stockhausen and other integral serialists of the Darmstadt School […] 27 Informaon Extrac0on: En0ty Linking Anton Webern (b Vienna, 3 Dec 1883; d Mi\ersill,15 Sept 1945). Austrian composer and conductor. Webern, who was probably Schoenberg's first private pupil, and Alban Berg, who came to him a few weeks later […] Boulez and Stockhausen and other integral serialists of the Darmstadt School […] Domain Knowledge Base 28 Informaon Extrac0on: Relaon Extrac0on • Webern, who was probably Schoenberg's first private pupil 29 Informaon Extrac0on: Relaon Extrac0on • Webern, who was probably Schoenberg's first private pupil Webern Schoenberg pupil_of 30 Methodology 31 Knowledge Graph 32 Knowledge Graph: Data Analy0cs • 16,707 biographies Role Amount • 434 roles composer 2618 teacher 1065 • Graph conductor 968 – 47,367 nodes pianist 704 – 274,333 edges organist 676 singer 404 violinist 285 … musicologist 144 cric 133 33 Birth Year Death Year 34 Birth Year Death Year 35 Knowledge Graph: Data Analy0cs Country Births Deaths Difference United States 2317 2094 -10% Italy 1616 1279 -21% Germany 1270 1292 2% France 991 1058 7% United Kingdom 882 877 -1% City Births Deaths Difference London 322 507 57% Paris 304 720 137% New York 266 501 88% Vienna 177 292 65% Rome 159 256 61% 36 Knowledge Graph: En0ty Relevance • City – Paris, London, Vienna, Rome, Venice, Berlin, Paris, New York • Venue – Covent Garden Theatre, King's Theatre, Drury Lane, Carnegie Hall, Théâtre de la Monnaie, Stad\heater, Theatre Royal, … • Educaonal Ins0tu0on – Paris Conservatoire, Moscow Conservatory, Juilliard School, St Petersburg Conservatory, Bmus, Prague Conservatory, Leipzig Conservatory, Vienna Hochschule für Musik, ... 37 Knowledge Graph: En0ty Relevance • Biography subject – Haydn, Claude Debussy, Arnold Schoenberg, Robert Stevenson, Paul Hindemith, Giovanni Pierluigi da Palestrina, Gustav Mahler, Maurice Ravel, Jean-Philippe Rameau – Mozart?, Bach?, Wagner? • Genre – chamber music, cappella, jazz, folk music, avant garde, baroque music, electronic music, musical theatre, plainchant 38 Knowledge Graph: En0ty Similarity • PageRank algorithm and Maximal Common Subgraph • Arnold Schoenberg: Anton Webern, Paul Hindemith, Gustav Mahler, Alban Berg, Claude Debussy • Guido Adler: Heinrich Jalowetz, Eusebius Mandyczewski, Robert Fuchs, Karl Weigl, Anton Wranitzky • Manuel de Falla: Ricardo Viñes, Juan Vicente Lecuna, Enrique Granados, Miguel Llobet Soles, Luigi Russolo • Miles Davis: Dizzy Gillespie, Herbie Hancock, Paul Chambers, Tony Williams, Cannonball Adderley 39 Relaons Graph: Informaon Visualizaon 40 Relaons Graph: Informaon Visualizaon 41 Mixing Different Data Sources • Creaon and curaon of a library from Web content • FlaBase: Flamenco Knowledge Base 42 Mixing Different Data Sources • Data Acquisi0on – APIs – Web crawling – SPARQL endpoints • En0ty Resolu0on – String similarity between labels – Graph similarity (context informaon) 43 FlaBase: Flamenco Knowledge Base 44 FlaBase: Flamenco Knowledge Base • Data gathered – 1,174 Ar0sts (text biography) – 76 Palos (flamenco genres) – 2,913 Albums – 14,078 Tracks – 771 Andalusian locaons • Knowledge Extracted – Place of birth – Date of birth – En0ty men0ons in text 45 FlaBase: Data Analy0cs • Number of ar0sts by year of birth 46 FlaBase: Data Analy0cs 47 FlaBase: Ar0st Relevance • Flamenco expert evaluaon Precision Values 48 Conclusions • Music Digital Libraries can benefit from seman0c approaches • Music Digital Libraries are s0ll in an early stage of development compared to the Web (Linked Open Data, Google Knowledge Graph) • Knowledge acquisi0on from Digital Libraries can help musicologists not only to search content, but also to discover new knowledge 49 Aknowledgments • This work was partly funded by the COFLA2 research project (Proyectos de Excelencia de la Junta de Andalucía, FEDER P12- TIC-1362). 50 Bibliography • Oramas S., Sordo M., Espinosa-Anke L., Serra X. (2015). A Seman,c-based approach for Ar,st Similarity. Internaonal Society for Music Informaon Retrieval Conference ISMIR 2015. In Press. • Oramas S., Gomez F., Gomez E., Mora J. (2015). FlaBase: Towards the crea,on of a Flamenco Music Knowledge Base. Internaonal Society for Music Informaon Retrieval Conference ISMIR 2015. In Press. • Sordo, M., Oramas S., & Espinosa-Anke L. (2015). Extracng Rela,ons from Unstructured Text Sources for Music Recommenda,on. Internaonal Conference on Applicaons of Natural Language to Informaon Systems NLDB 2015. • Oramas S., Sordo M., Espinosa-Anke L. (2015). A Rule-based Approach to Extracng Rela,ons from Music Tidbits. 2nd Workshop on Knowledge Extrac0on from Text at WWW 2015. • Oramas S., Sordo M., Serra X. (2014). Automa,c Crea,on of Knowledge Graphs from Digital Musical Document Libraries. Conference in Interdisciplinary Musicology CIM 2014. 51 Thanks! Knolwedge Acquisi0on from Music Digital Libraries Sergio Oramas, Mohamed Sordo [email protected] @sergiooramas .
Load more

Recommended publications
  • 17 by Francis Mallett and Simon Obert on 20 June 1962 Johanna
    Beiträge A Case of Mistaken Identity? The Jalowetz Portrait by Richard Gerstl by Francis Mallett and Simon Obert On 20 June 1962 Johanna Jalowetz, the widow of the conductor and Schoenberg pupil Heinrich Jalowetz (d. 1946), wrote a letter to the art historian Otto Breicha. In those years Breicha was conducting research into the painter Richard Gerstl, in particular gathering information for a cata- logue of his works. In this connection he turned to Johanna Jalowetz, knowing that she owned “a painting by the artist.”1 Johanna Jalowetz re- sponded to his questions by noting that she possessed a Gerstl self-portrait, a “pointillist drawing” in ink on paper that probably “dates from the year 1906/7.” It bore, she continued, an “extraordinary resemblance” to the artist and was “a gift from him.”2 But she then added a paragraph whose contents can only have taken Breicha by surprise: At the same time, Gerstl painted to order a life-size three-quarter-length oil portrait of my husband in a pointillist technique. When we moved away from Vienna we left the said portrait for safekeeping with a relative, Alois Kurzweil. Our many changes of residence (at that time an opera conductor’s lot) made it impossible for us to look after the painting. And by the time we became more settled, after the First World War, we had lost all contact with and trace of Alois Kurzweil.3 1 Otto Breicha, letter of 4 June 1962 to Johanna Jalowetz, carbon copy in the Otto Breicha Archive, Leopold Museum, Vienna. We wish to thank Dominik Papst of the Leopold Museum for making available to us this and the following letter (see note 3).
    [Show full text]
  • German Jews in the United States: a Guide to Archival Collections
    GERMAN HISTORICAL INSTITUTE,WASHINGTON,DC REFERENCE GUIDE 24 GERMAN JEWS IN THE UNITED STATES: AGUIDE TO ARCHIVAL COLLECTIONS Contents INTRODUCTION &ACKNOWLEDGMENTS 1 ABOUT THE EDITOR 6 ARCHIVAL COLLECTIONS (arranged alphabetically by state and then city) ALABAMA Montgomery 1. Alabama Department of Archives and History ................................ 7 ARIZONA Phoenix 2. Arizona Jewish Historical Society ........................................................ 8 ARKANSAS Little Rock 3. Arkansas History Commission and State Archives .......................... 9 CALIFORNIA Berkeley 4. University of California, Berkeley: Bancroft Library, Archives .................................................................................................. 10 5. Judah L. Mages Museum: Western Jewish History Center ........... 14 Beverly Hills 6. Acad. of Motion Picture Arts and Sciences: Margaret Herrick Library, Special Coll. ............................................................................ 16 Davis 7. University of California at Davis: Shields Library, Special Collections and Archives ..................................................................... 16 Long Beach 8. California State Library, Long Beach: Special Collections ............. 17 Los Angeles 9. John F. Kennedy Memorial Library: Special Collections ...............18 10. UCLA Film and Television Archive .................................................. 18 11. USC: Doheny Memorial Library, Lion Feuchtwanger Archive ...................................................................................................
    [Show full text]
  • A Heretic in the Schoenberg Circle: Roberto Gerhard's First Engagement with Twelve-Tone Procedures in Andantino
    Twentieth-Century Music 16/3, 557–588 © Cambridge University Press 2019. This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution, and reproduction in any medium, provided the original work is properly cited. doi: 10.1017/S1478572219000306 A Heretic in the Schoenberg Circle: Roberto Gerhard’s First Engagement with Twelve-Tone Procedures in Andantino DIEGO ALONSO TOMÁS Abstract Shortly before finishing his studies with Arnold Schoenberg, Roberto Gerhard composed Andantino,a short piece in which he used for the first time a compositional technique for the systematic circu- lation of all pitch classes in both the melodic and the harmonic dimensions of the music. He mod- elled this technique on the tri-tetrachordal procedure in Schoenberg’s Prelude from the Suite for Piano, Op. 25 but, unlike his teacher, Gerhard treated the tetrachords as internally unordered pitch-class collections. This decision was possibly encouraged by his exposure from the mid- 1920s onwards to Josef Matthias Hauer’s writings on ‘trope theory’. Although rarely discussed by scholars, Andantino occupies a special place in Gerhard’s creative output for being his first attempt at ‘twelve-tone composition’ and foreshadowing the permutation techniques that would become a distinctive feature of his later serial compositions. This article analyses Andantino within the context of the early history of twelve-tone music and theory. How well I do remember our Berlin days, what a couple we made, you and I; you (at that time) the anti-Schoenberguian [sic], or the very reluctant Schoenberguian, and I, the non-conformist, or the Schoenberguian malgré moi.
    [Show full text]
  • Verklärte Nacht Works for Piano Trio Osiris Trio
    ARNOLD SCHÖNBERG / KARL WEIGL Verklärte Nacht Works for Piano Trio Osiris Trio 1 ARNOLD SCHÖNBERG / KARL WEIGL Verklärte Nacht Works for Piano Trio Osiris Trio Ellen Corver piano Peter Brunt violin Larissa Groeneveld cello ARNOLD SCHÖNBERG (1874-1951) In September 1899, Arnold history. Schönberg’s starting point [1] Verklärte Nacht op. 4 (1899) 31:46 Schönberg was on vacation in the was one of the poems from Dehmel’s Arranged for piano trio by Eduard Steuermann Austrian village of Payerbach am Weib und Welt cycle, which raised Semmering with his friend, Alexander a furore following its publication in KARL WEIGL (1881-1949) von Zemlinsky, and Alexander’s sister 1896: conservative critics found the Piano Trio (1938-39) Mathilde. Schönberg was head over poems to be erotic, immoral and even [2] Allegro moderato 9:47 heels in love with Mathilde (they blasphemous. [3] Andante 8:27 were married 2 years later) and had [4] Allegro molto 5:09 also fallen completely under the Schönberg had already composed spell of the work of the German poet some songs to poems by Dehmel in Richard Dehmel – famous but at 1897, but it was not until 1899 – his total time 55:10 the same time highly controversial, ‘Dehmel year’ – that he felt fully up fêted by some and vilified by others. to the task of setting the violent Schönberg’s own development as emotions of this poetry to music in a a composer was rapid indeed, as convincing way: as well as Verklärte he successfully transformed the Nacht, he set about working on eight influences of Brahms and Wagner songs based on Dehmel’s poems.
    [Show full text]
  • INFORMATION to USERS This Manuscript Has Been Reproduced
    INFORMATION TO USERS This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer. The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book. Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. UMI A Bell & Howell Information Company 300 North Zeeb Road. Ann Arbor. Ml 48106-1346 USA 313/761-4700 800/521-0600 THE COMPLETED SYMPHONIC COMPOSITIONS OF ALEXANDER ZEMLINSKY DISSERTATION Volume I Presented in Partial Fulfillment of the Requirement for the Degree Doctor of Philosophy In the Graduate School of The Ohio State University By Robert L.
    [Show full text]
  • Semantic Search
    Semantic Search Philippe Cudre-Mauroux Definition ter grasp the semantics (i.e., meaning) and the context of the user query and/or Semantic Search regroups a set of of the indexed content in order to re- techniques designed to improve tra- trieve more meaningful results. ditional document or knowledge base Semantic Search techniques can search. Semantic Search aims at better be broadly categorized into two main grasping the context and the semantics groups depending on the target content: of the user query and/or of the indexed • techniques improving the relevance content by leveraging Natural Language of classical search engines where the Processing, Semantic Web and Machine query consists of natural language Learning techniques to retrieve more text (e.g., a list of keywords) and relevant results from a search engine. results are a ranked list of documents (e.g., webpages); • techniques retrieving semi-structured Overview data (e.g., entities or RDF triples) from a knowledge base (e.g., a knowledge graph or an ontology) Semantic Search is an umbrella term re- given a user query formulated either grouping various techniques for retriev- as natural language text or using ing more relevant content from a search a declarative query language like engine. Traditional search techniques fo- SPARQL. cus on ranking documents based on a set of keywords appearing both in the user’s Those two groups are described in query and in the indexed content. Se- more detail in the following section. For mantic Search, instead, attempts to bet- each group, a wide variety of techniques 1 2 Philippe Cudre-Mauroux have been proposed, ranging from Natu- matical tags (such as noun, conjunction ral Language Processing (to better grasp or verb) to individual words.
    [Show full text]
  • SEKI@Home, Or Crowdsourcing an Open Knowledge Graph
    SEKI@home, or Crowdsourcing an Open Knowledge Graph Thomas Steiner1? and Stefan Mirea2 1 Universitat Politècnica de Catalunya – Department LSI, Barcelona, Spain [email protected] 2 Computer Science, Jacobs University Bremen, Germany [email protected] Abstract. In May 2012, the Web search engine Google has introduced the so-called Knowledge Graph, a graph that understands real-world en- tities and their relationships to one another. It currently contains more than 500 million objects, as well as more than 3.5 billion facts about and relationships between these different objects. Soon after its announce- ment, people started to ask for a programmatic method to access the data in the Knowledge Graph, however, as of today, Google does not provide one. With SEKI@home, which stands for Search for Embedded Knowledge Items,weproposeabrowserextension-basedapproachtocrowdsource the task of populating a data store to build an Open Knowledge Graph. As people with the extension installed search on Google.com, the ex- tension sends extracted anonymous Knowledge Graph facts from Search Engine Results Pages (SERPs) to a centralized, publicly accessible triple store, and thus over time creates a SPARQL-queryable Open Knowledge Graph. We have implemented and made available a prototype browser extension tailored to the Google Knowledge Graph, however, note that the concept of SEKI@home is generalizable for other knowledge bases. 1Introduction 1.1 The Google Knowledge Graph With the introduction of the Knowledge Graph, the search engine Google has made a significant paradigm shift towards “things, not strings” [7], as a post on the official Google blog states. Entities covered by the Knowledge Graph include landmarks, celebrities, cities, sports teams, buildings, movies, celestial objects, works of art, and more.
    [Show full text]
  • Walter Frisch, Ed. Brahms and His World, Princeton: Princeton University Press, 1990
    Document generated on 09/23/2021 8:36 a.m. Canadian University Music Review Revue de musique des universités canadiennes --> See the erratum for this article Walter Frisch, ed. Brahms and His World, Princeton: Princeton University Press, 1990. 223 pp. ISBN 0-691-09139-0 (cloth), ISBN 0-691-02713-7 (paper) James Deaville Volume 12, Number 1, 1992 URI: https://id.erudit.org/iderudit/1014221ar DOI: https://doi.org/10.7202/1014221ar See table of contents Publisher(s) Canadian University Music Society / Société de musique des universités canadiennes ISSN 0710-0353 (print) 2291-2436 (digital) Explore this journal Cite this review Deaville, J. (1992). Review of [Walter Frisch, ed. Brahms and His World, Princeton: Princeton University Press, 1990. 223 pp. ISBN 0-691-09139-0 (cloth), ISBN 0-691-02713-7 (paper)]. Canadian University Music Review / Revue de musique des universités canadiennes, 12(1), 149–154. https://doi.org/10.7202/1014221ar All Rights Reserved © Canadian University Music Society / Société de musique This document is protected by copyright law. Use of the services of Érudit des universités canadiennes, 1991 (including reproduction) is subject to its terms and conditions, which can be viewed online. https://apropos.erudit.org/en/users/policy-on-use/ This article is disseminated and preserved by Érudit. Érudit is a non-profit inter-university consortium of the Université de Montréal, Université Laval, and the Université du Québec à Montréal. Its mission is to promote and disseminate research. https://www.erudit.org/en/ was not a first language. Descriptive accounts and, even worse, abstract ideas in several papers (especially those by Czechanowska, Reinhard, and Yeh) are unclear, seemingly muddied by the process of translation.
    [Show full text]
  • Seo Report for Brands July 2019
    VOICE ASSISTANT SEO REPORT FOR BRANDS JULY 2019 GIVING VOICE TO A REVOLUTION Table of Contents About Voicebot About Magic + Co An Introduction to Voice Search Today // 3 Voicebot produces the leading independent Magic + Co. is a next-gen creative services firm research, online publication, newsletter and podcast whose core expertise is in designing, producing, and Consumer Adoption Trends // 10 focused on the voice and AI industries. Thousands executing on conversational strategy, the technology of entrepreneurs, developers, investors, analysts behind it, and associated marketing campaigns. Voice Search Results Analysis // 16 and other industry leaders look to Voicebot each Grading the Experts // 32 week for the latest news, data, analysis and insights Facts defining the trajectory of the next great computing • 25 people, technologists, strategists, and Practical Voice Assistant SEO Strategies // 36 platform. At Voicebot, we give voice to a revolution. creatives with focus on conversational design and technologies. Additional Resources // 39 • 1 Webby Award Nominee and 3 Honorees. • 3 years old and global leader. Methodology • We’ve worked primarily with CPG, but our practice extends into utilities, governments, Consumer survey data was collected online during banking, entertainment, healthcare and is the first week of January 2019 and included 1,038 international in scope. U.S. adults age 18 or older that were representative of U.S. Census demographic averages. Findings were compared to earlier survey data collected in Magic + Co. the first week of September 2018 that included responses from 1,040 U.S. adults. Voice search data results from Amazon Echo Smart Speakers with Alexa, Apple HomePod with Siri, Google Home and a Smartphone with Google Assistant, and Samsung Galaxy S10 with Bixby, were collected in Q2 2019.
    [Show full text]
  • “Tragic” Mistake
    UNDOING A “TRAGIC” MISTAKE DETERMINING THE INNER-MOVEMENT ORDER OF MAHLER’S SIXTH SYMPHONY A critical examination of the evidence by Jerry Bruck New York City October 19th, 2002 A publication of THE KAPLAN FOUNDATION 450 Park Avenue New York City © Jerry Bruck, 2002 I. OVERVIEW Nearly a century has passed since Gustav Mahler composed his Sixth Symphony, yet confusion still persists among conductors, scholars and biographers regarding the order of its inner movements. Mahler began work on the symphony in 1903, first composing a Scherzo and an Andante as the central pair of its eventual four-movement structure, framing them with the remaining movements the following year. He then reversed this “S-A” order of inner movements before the symphony’s premiere in 1906, and thereafter never reverted to their previous arrangement. It was not until 1919, almost a decade after Mahler’s death, that the conductor Willem Mengelberg queried Mahler’s widow about the order of these inner movements. Her response: “First Scherzo, then Andante” prompted him to alter the “A-S” order of his conductor’s score, igniting a controversy that has spanned the decades since. With the publication in 1963 of the first Critical Edition of the Sixth by the Internationale Gustav Mahler Gesellschaft (IGMG), the matter seemed settled at last. In his introduction, IGMG founder-editor Erwin Ratz stated that the thematic similarities between the symphony’s opening movement and its Scherzo, commented upon during rehearsals for its premiere, had prompted Mahler to succumb to the advice of “outside influences” to transpose the Sixth’s inner movements.
    [Show full text]
  • Arxiv:1812.02903V1 [Cs.LG] 7 Dec 2018 Ized Server
    APPLIED FEDERATED LEARNING: IMPROVING GOOGLE KEYBOARD QUERY SUGGESTIONS Timothy Yang*, Galen Andrew*, Hubert Eichner* Haicheng Sun, Wei Li, Nicholas Kong, Daniel Ramage, Franc¸oise Beaufays Google LLC, Mountain View, CA, U.S.A. ftimyang, galenandrew, huberte haicsun, liweithu, kongn, dramage, [email protected] ABSTRACT timely suggestions are necessary in order to maintain rele- vance. On-device inference and training through FL enable Federated learning is a distributed form of machine learn- us to both minimize latency and maximize privacy. ing where both the training data and model training are decen- In this paper, we use FL in a commercial, global-scale tralized. In this paper, we use federated learning in a commer- setting to train and deploy a model to production for infer- cial, global-scale setting to train, evaluate and deploy a model ence – all without access to the underlying user data. Our to improve virtual keyboard search suggestion quality with- use case is search query suggestions [4]: when a user enters out direct access to the underlying user data. We describe our text, Gboard uses a baseline model to determine and possibly observations in federated training, compare metrics to live de- surface search suggestions relevant to the input. For instance, ployments, and present resulting quality increases. In whole, typing “Let’s eat at Charlie’s” may display a web query sug- we demonstrate how federated learning can be applied end- gestion to search for nearby restaurants of that name; other to-end to both improve user experiences and enhance user types of suggestions include GIFs and Stickers. Here, we privacy.
    [Show full text]
  • Alban Berg's Filmic Music
    Louisiana State University LSU Digital Commons LSU Doctoral Dissertations Graduate School 2002 Alban Berg's filmic music: intentions and extensions of the Film Music Interlude in the Opera Lula Melissa Ursula Dawn Goldsmith Louisiana State University and Agricultural and Mechanical College, [email protected] Follow this and additional works at: https://digitalcommons.lsu.edu/gradschool_dissertations Part of the Music Commons Recommended Citation Goldsmith, Melissa Ursula Dawn, "Alban Berg's filmic music: intentions and extensions of the Film Music Interlude in the Opera Lula" (2002). LSU Doctoral Dissertations. 2351. https://digitalcommons.lsu.edu/gradschool_dissertations/2351 This Dissertation is brought to you for free and open access by the Graduate School at LSU Digital Commons. It has been accepted for inclusion in LSU Doctoral Dissertations by an authorized graduate school editor of LSU Digital Commons. For more information, please [email protected]. ALBAN BERG’S FILMIC MUSIC: INTENTIONS AND EXTENSIONS OF THE FILM MUSIC INTERLUDE IN THE OPERA LULU A Dissertation Submitted to the Graduate Faculty of the Louisiana State University and Agricultural and Mechanical College in partial fulfillment of the requirements for the degree of Doctor of Philosophy in The College of Music and Dramatic Arts by Melissa Ursula Dawn Goldsmith A.B., Smith College, 1993 A.M., Smith College, 1995 M.L.I.S., Louisiana State University and Agricultural and Mechanical College, 1999 May 2002 ©Copyright 2002 Melissa Ursula Dawn Goldsmith All rights reserved ii ACKNOWLEDGMENTS It is my pleasure to express gratitude to my wonderful committee for working so well together and for their suggestions and encouragement. I am especially grateful to Jan Herlinger, my dissertation advisor, for his insightful guidance, care and precision in editing my written prose and translations, and open mindedness.
    [Show full text]