The International Conference on Computational Linguistics and Intelligent Text Processing (CICLing) is an annual conference on computational linguistics (CL) and natural language processing (NLP). The conference is intended to encourage exchange of opinions between scientists working in different areas related to computational linguistics, intelligent text and speech processing.

The 20th edition of CICLing took place in La Rochelle during 7-14 April, where Lorenzo Porcaro (from UPF) presented his work on named entity recognition in user-generated content focusing on the music domain, done in collaboration with Horacio Saggion (TALN Natural Language Processing Group, Universitat Pompeu Fabra).

 

Lorenzo Porcaro presenting at CICLing 2019

 

Recognizing musical entities is important for Music Information Retrieval (MIR), since it can improve the performance of several tasks such as music recommendation, genre classification or artist similarity. However, most entity recognition systems in the music domain have concentrated on formal texts (e.g. artists’ biographies, encyclopedic articles, etc.), ignoring rich and noisy user-generated content.

This study analyzes a set of tweets related to a classical music radio channel, interested in detecting two types of entities: Contributor (composer, performer, conductor, etc.) and Musical Work (symphony, concerto, overture, etc.).



Reference

Lorenzo Porcaro, Horacio Saggion (2019). Recognizing Musical Entities in User-generated Content.  International Conference on Computational Linguistics and Intelligent Text Processing (CICLing) 2019, University of La Rochelle, La Rochelle, 7-13 April.