Implementasi Algoritma Topic Modeling pada Abstrak Paper Ilmiah untuk Deteksi Tren Topik Penelitian

Understanding the development and research focus in the field of language, as well as evaluating language resources, are crucial to support further research endeavors. This study aims to detect research topic trends using the Topic Modeling algorithm on abstracts of scientific papers. The method applied is Latent Dirichlet Allocation (LDA), enabling automatic topic identification from a document corpus. The research process involves several stages, including data collection, preprocessing, topic weighting, and analysis of LDA model outcomes. The LDA model is then utilized to cluster data into dominant topics. The findings indicate the effectiveness of the LDA method in identifying research topic trends. In the LDA model with 3 topics, it reflects the complexity of natural language processing, from technical aspects to efforts in resource development and linguistic research. Meanwhile, in the LDA model with 5 topics, it provides an overview of text and language analysis. This research is anticipated to contribute to the fields of text analysis and topic modeling, aiding researchers in identifying recent trends in scientific literature.