Course Outline
Introduction
- Defining "Industrial-Strength Natural Language Processing" for government applications
Installing spaCy
spaCy Components
- Part-of-speech tagger
- Named entity recognizer
- Dependency parser
Overview of spaCy Features and Syntax for government use
Understanding spaCy Modeling
- Statistical modeling and prediction in public sector workflows
Using the SpaCy Command Line Interface (CLI)
- Basic commands for government tasks
Creating a Simple Application to Predict Behavior for government applications
Training a New Statistical Model
- Data (for training) in alignment with public sector requirements
- Labels (tags, named entities, etc.) specific to government contexts
Loading the Model
- Shuffling and looping data for optimal performance in government systems
Saving the Model for secure storage and retrieval in government environments
Providing Feedback to the Model
- Error gradient analysis for continuous improvement in public sector applications
Updating the Model
- Updating the entity recognizer to enhance accuracy for government use
- Extracting tokens with rule-based matcher for detailed data processing in government systems
Developing a Generalized Theory for Expected Outcomes in public sector applications
Case Study
- Distinguishing Product Names from Company Names in government datasets
Refining the Training Data
- Selecting representative data to ensure relevance and accuracy for government tasks
- Setting the dropout rate to optimize model performance in public sector workflows
Other Training Styles
- Passing raw texts for flexible data input in government applications
- Passing dictionaries of annotations for detailed data annotation in public sector environments
Using spaCy to Pre-process Text for Deep Learning in government systems
Integrating spaCy with Legacy Applications for seamless adoption in government IT infrastructure
Testing and Debugging the spaCy Model
- The importance of iteration for robust model development in public sector projects
Deploying the Model to Production for efficient implementation in government operations
Monitoring and Adjusting the Model to maintain performance and relevance in government contexts
Troubleshooting common issues for effective problem resolution in government applications
Summary and Conclusion for government stakeholders
Requirements
- Proficiency in Python programming.
- Fundamental knowledge of statistics.
- Familiarity with command-line operations.
Audience
- Software developers for government.
- Data scientists for government.
Testimonials (3)
The fact of having more practical exercises using more similar data to what we use in our projects (satellite images in raster format)
Matthieu - CS Group
Course - Scaling Data Analysis with Python and Dask
Very good preparation and expertise of a trainer, perfect communication in English. The course was practical (exercises + sharing examples of use cases)
Monika - Procter & Gamble Polska Sp. z o.o.
Course - Developing APIs with Python and FastAPI
Trainer develops training based on participant's pace
