Media portrayals of libraries in Indonesia: A large-language-model-assisted content analysis

Libraries often occupy an under-represented role in public discourse, particularly in developing countries such as Indonesia. In the digital era, public understanding of libraries is partly shaped by news media, yet limited scholarship has examined how digital outlets portray them. This study analyzes coverage of libraries on two major Indonesian news platforms—Detik.com and Kompas.com—focusing on news topics, news tones, news types, library roles, and news sources. A quantitative content analysis was conducted on 166 articles published in 2022–2023. Coding was undertaken by both a human coder and a large language model during the initial reliability testing of the coding protocol, after which full-scale coding was performed with the large language model. Statistical analyses were applied to assess significant differences. The results show that Kompas.com emphasizes formal, policy-oriented reporting, while Detik.com highlights user-focused narratives. These differences suggest that library stakeholders should adapt their communication strategies to each outlet's editorial style.