Combining NLP with Structured Data
summary
Claims a practical hybrid NLP+structured-data pipeline implemented in Python that links extracted clinical entities to document section headers. Describes concrete components: text normalization, sentence classification and sequence labeling, use of token/sequence embeddings plus structured metadata for disambiguation, and heuristics for noisy layouts. Presents an end-to-end mapper applied to real clinical documents, reports qualitative error modes (layout, abbreviations) and deployment lessons showing the approach is workable but needs task-specific tuning.
AI-written summary · about
chapters (5)
- 0:00Introduction
- 3:52Medical coding basics
- 17:19Sentence classification approach
- 23:40Hybrid NLP model
- 30:43Q&A and conclusion
related talks
26mNatural Language Processing for Healthcare: Current Trends...
06mAdvancements in Natural Language Processing (NLP) and Text Analytics
34mExposing Dark Data in the enterprise with custom NLP
27mMarrying Structured + Unstructured Data
44mExtracting Structured Data from Legal Documents
26mNatural Language Processing (NLP) going mainstream
38mMaking neural content meaningful and truthful relying on OpenNLP tools
16mGenerative AI for Claims and Fraud