<![CDATA[People's perceptions of orangutans strongly influence their responses in every interaction, which can have negative impacts on both humans and orangutans. Misperceptions, fear, or negative attitudes may increase the risk of conflict, injury, or retaliatory actions, thereby threatening conservation outcomes. Understanding community sentiment toward human–orangutan interactions is therefore essential for designing effective conservation strategies. However, direct data collection to understand community sentiment often faces cost and accuracy constraints. Therefore, an alternative approach that is more efficient and objective is needed to capture community sentiment toward human-orangutan interactions. This research offers latent topic modelling and sentiment analysis of online news articles as a novel method to understand these dynamics and support conservation efforts. A total of 11 latent topics were obtained from the news articles. Orangutan sightings and handovers of orangutans were the most frequently discussed in 2016. Palm oil plantations emerged as a prominent topic related to human-orangutan interaction incidents. Negative sentiment was predominantly linked to topic such as injured orangutans and orangutan sightings on plantations, whereas orangutan translocation received the highest positive sentiment score. This study highlights the potential of natural language processing for analyzing Indonesian language texts in conservation contexts, with applications extendable to broader environmental and forestry issues such as deforestation and wildlife hunting.]]>
Copyrights © 2026
