Topic modelling of diverse events in the 21st century public library: A case analysis of New York Public Library

Abstract

Public libraries hold events to meet the needs of citizens. Recently, online events have increased due to COVID-19. The purpose of this study is to describe public library events with an emphasis on those held during COVID-19. Analysis was based on 21,411 data points collected from event descriptions on the New York Public Library (NYPL) website, from November 2020 to November 2021. Using Latent Dirichlet Allocation (LDA), a topic modelling method, data were collected and analysed. Analysis. Event topics were extracted from event descriptions using LDA. 24 event topics were identified. Five general topical categories were then derived qualitatively from the 24 topics. The topics and topical categories were visualised using PCoA, a general algorithm used in LDA. 24 topics and 5 general topical categories were created. The events ranged from traditional collection-based events to future-oriented events to help overcome community division. Online participation was a common theme among all events. Online events that developed due to the COVID-19 pandemic offered increased opportunities for participation. In the current uncertain climate, public library events play an important role in supporting everyday life and culture.