In this workshop, we will focus on hands-on activities that simplify the end-to-end experience of developing custom models for Text Mining in Azure Machine Learning.
No prior experience with Text Mining is assumed, although knowledge of Natural Language Processing or Computational Linguistics will be helpful. These labs assume an intermediate knowledge of Azure Machine Learning Workbench, and if this is not the case, then you should spend the time working through the pre-requisites.
Pre-requisites can be found here. Briefly, pre-requisites include the following:
- The ability to create resources within an Azure subscription
- Familiarity with how to create resources in said subscription
Azure Machine Learning Workbench
- Python Proficiency and familiarity with Azure Machine Learning Workbench’s end-end features including deployment
AML Package for Text Analytics
- To download and install the packages, follow the instructions in the pre-requisites
Azure ML Package for Text Analytics (TATK)
- Understand how to develop a pipeline using TATK for text classification
- Understand how to develop a pipline consisting of CNNs together with a pre-trained embedding model
- Build custom embeddings (word2vec and fastText)
- Understand how to train a custom entity extraction model
- Deploy trained models to Azure
Please note: This is a rough agenda, and the schedule is subject to change pending class activities and interaction.
- 9:00 - 10:00: Introduction and Context
- 10:00 - 11:00: Data Ingestion and Text Classification
- 11:00 - 12:00: Sentiment Analysis using CNNs and pre-trained word2vec model
- 12:00 - 1:00: Lunch
- 1:00 - 2:00: Custom Embeddings (word2Vec and fastText)
- 2:00 - 3:00: Entity Extraction
- 3:00 - 4:00: Deployment of pipeline
- 4:00 - 5:00: Q&A and Feedback