You are working on an NLP system that needs to extract structure from unstructured text or discover themes in a document collection. Depending on the use case, you may need a supervised sequence labeling approach or an unsupervised topic discovery workflow.
How do you approach named entity recognition or topic modeling?