Hao defends dissertation

Doctoral candidate Yun Hao successfully defended her dissertation, "Towards a Better Understanding of Music Playlist Titles and Descriptions," on April 6.

Her committee included Professor J. Stephen Downie (chair); Associate Professor Vetle Torvik; Assistant Professor Nigel Bosch; Xiao Hu, associate professor at The University of Hong Kong; and Andreas Ehmann, manager of research/data science at Pandora Inc.

Abstract: Music playlists, either user-generated or curated by music streaming services, often come with titles and descriptions. Although informative, these titles and descriptions make up a sparse and noisy semantic space that is challenging to be leveraged for tasks such as making music recommendations. This dissertation is dedicated to developing a better understanding of playlist titles and descriptions by leveraging track sequences in playlists. Specifically, work has been done to capture latent patterns in tracks by an embedding approach, and the latent patterns are found to be well aligned with the organizing principles of mix tapes identified more than a decade ago. Effectiveness of the latent patterns is evaluated by the task of generating descriptive keywords/tags for playlists given tracks, indicating that the latent patterns learned from tracks in playlists are able to provide a good understanding of playlist titles and descriptions. The identified latent patterns are further leveraged to improve model performance on the task of predicting missing tracks given playlist titles and descriptions. Experimental results show that the proposed models yield improvements to the task, especially when playlist descriptions are provided as model input in addition to titles. Main contributions of this work include (1) providing a better solution to dealing with "cold-start'' playlists in music recommender systems, and (2) proposing an effective approach to automatically generating descriptive keywords/tags for playlists using track sequences.

Updated on
Backto the news archive

Related News

Cheng defends dissertation

Doctoral candidate Jessica Cheng successfully defended her dissertation, "Agreeing to Disagree: Applying a Logic-based Approach to Reconciling and Merging Multiple Taxonomies," on May 25. 

Jessica Cheng

Student award recipients announced

Each year, the School of Information Sciences recognizes a group of outstanding students for their achievement in academics as well as a number of attributes that contribute to professional success. Congratulations to this year's honorees!

Alma with cap

Brooks presents keynote at West African conference

Ian Brooks, iSchool research scientist and director of the Center for Health Informatics (CHI), gave a keynote talk at the West Africa Conference on Digital Public Goods and Cybersecurity, which was held on May 9-10 in Freetown, Sierra Leone. The conference focused on bridging the gender gap in digital public goods and cybersecurity spaces in Africa.

Ian Brooks

New project to help identify and predict insider threats

Insider threats are one of the top security concerns facing large organizations. Current and former employees, business partners, contractors—anyone with the right level of access to a company’s data—can pose a threat. The incidence of insider threats has increased in recent years, at a significant cost to companies. Associate Professor Jingrui He is addressing this problem in a new project that seeks to detect and predict insider threats. She has been awarded a three-year, $200,000 grant from the C3.ai Digital Transformation Institute for her project, "Multi-Facet Rare Event Modeling of Adaptive Insider Threats."

Jingrui He

iSchool students present their research at Urbana City Council meeting

At the Urbana City Council meeting on May 9, students in the Community Data (IS 594) course presented their research on how communities are reducing gun violence. According to their instructor Chamee Yang, postdoctoral research associate with the iSchool, Community Data Clinic, and Just Infrastructures Initiative, the new course was designed as an experiential learning opportunity with a community engagement component, where students could gain research experience with real-world implications. Throughout the Spring 2022 semester, students worked in groups to explore community-driven approaches to prevent gun violence.

Chamee Yang, Sarah Unruh, and Gowri Balasubramaniam