Hybrid LLM-enhanced topic modelling for large-scale thematic analysis of dairy cattle health literature

The dataset contains the outputs of BERTopic algorithm. This is a unsupervised clustering tool used to group unstructured bodies of text. In this research project, the texts are scientific abstracts related to a give domain of research. The latter concerns Dairy Cattle Health. BERTopic's outputs have been enhanced with a LLM deployed locally on the Ubelix HPC. This helped to summarize and labeled the topics and titles. In the dataset, you will find the BERTopic topic titltes (Name), the numero of the topic (Topic), sorted by the size of the cluster (Count). (Representation) highlights the main keywords found in the cluster to create the BERTopic's title. All other columns are created by LLamA LLM to help us understand better the topics and their content.