Content summary: Topic modeling involves extracting the most salient topics from large text corpora, such as collections of notes or articles. Traditional topic modeling methods, like Latent Dirichlet Allocation (LDA) and Structural Topic Modeling (STM), often fail to account for the context in which words appear. Advances in large language models (LLMs), capable of contextually understanding textual data, have led to the innovation of context-aware topic modeling techniques, such as BERTopic and TopClus. In this talk, we will discuss and demonstrate these advanced topic modeling techniques. The objectives include:
1. Explaining what topic modeling is and how it can be applied in our research.
2. Discussing how LLM-powered topic modeling techniques, like BERTopic and TopClus, differ from traditional methods, particularly in their use of pretrained LLMs to generate context-aware topics.
3. Providing a live demonstration of these techniques.
Presenter: Charles Alba
Code and materials used in this video can be downloaded from GitHub:
240127_BERTopic.zip; 240127_bertopic.pdf
github.com/Dre...
Hashtags: #topicmodeling #artificialintelligence #machinelearning #deeplearning #python #pythonprogramming #pythontutorial #aitutorial #coding #neuralnetworks #neuralnetwork #pytorch #computervision #nlp #naturallanguageprocessing #scikitlearn
9 сен 2024
