Language Technologies

Spotify listeners use language to express their needs, whether typing queries or speaking the names of songs they would like to hear. Additionally, songs and podcasts contain language that we can understand, classify, and match to user interests. We conduct research on all aspects of language technologies that are applicable to audio streaming. We help Spotify understand you, by being conversational, multilingual and interactive. We also learn the semantics of audio content and creators from language descriptions, including a knowledge graph of entities, ensuring our methods are scalable and include approaches to developing and maintaining shared vocabularies and ontologies. Our research areas range from computational linguistics, natural language processing and speech applications, to machine learning applied to all aspects of language.

Latest Language Technologies Publications

April 2021 | EACL

Detecting Extraneous Content in Podcasts

Sravana Reddy, Yongze Yu, Aasish Pappu, Aswin Sivaraman, Rezvaneh Rezapour, Rosie Jones

December 2020 | COLING

100,000 Podcasts: A Spoken English Document Corpus

Ann Clifton, Sravana Reddy, Yongze Yu, Aasish Pappu, Rezvaneh Rezapour, Hamed Bonab, Maria Eskevich, Gareth Jones, Jussi Karlgren, Ben Carterette, Rosie Jones

October 2020 | ISMIR - International Society for Music Information Retrieval Conference

“Play Music”: User Motivations and Expectations for Non-specific Voice Queries

Jennifer Thom, Angela Nazarian, Ruth Brillman, Henriette Cramer, Sarah Mennicken