Crash Course in Machine Learning for Natural Language Processing
summary
Shows how web text and images become structured info using OCR and Azure Form Recognizer, plus POS tagging, NER and vector embeddings to turn words into numbers. Demonstrates concrete pipelines: extract invoice/receipt fields (including handwriting), detect image/text mismatches for data quality, mine reviews for aspect-specific defects and product ideas, and standardize attributes across sources. Explains analogy math on embeddings for similarity searches and how multimodal pipelines combine extraction and proximity search for practical NLP systems.
AI-written summary · about
chapters (32)
- 0:00Intro
- 0:49Web Data
- 1:19Twitter
- 1:32YouTube
- 2:22AI vs Machine Learning
- 3:27AI is not Rocket Science
- 3:42Small Example
- 4:10Anomaly Detection
- 6:05Agenda
- 6:56Machine Learning
- 11:32Data vs Information
- 12:42Multimodal Analysis
- 13:58Sentiment Analysis
- 14:17Data Quality
- 16:39Review Analysis
- 18:30Review Mining Research
- 20:47Product Listings
- 22:23Language Manipulation
- 22:52Ambiguity
- 24:44Tools and Techniques
- 25:10Relationship and Context
- 25:53Grammar
- 26:33Tag Set
- 27:35Named Entity Recognition
- 28:24Words as Numbers
- 30:37Proximity
- 31:33Expanding
- 32:16Bring it all together
- 33:12multimodal
- 33:29multimodal tools
- 33:43books
- 35:05questions
related talks
06mAdvancements in Natural Language Processing (NLP) and Text Analytics
55mDemystifying Natural Language Processing across several industry verticals
48mIntroduction to Artificial Intelligence, Machine Learning, Deep Learning and Others
44mTeaching Computers ABCs: A Quick Intro to Natural Language Processing
31mDeep Learning Techniques for Natural Language Processing
33mApplied AI: Real-world use cases for Microsoft's Azure Cognitive Services
01h 39mNatural Language Processing (NLP) in Python - From Zero to Hero
47mBuilding smarter apps with Machine Learning, from magic to reality