Spark NLP Pipelines Training in Copenhagen, Denmark

This training provides a practical and project-based introduction to Spark NLP, a production-grade natural language processing library built on Apache Spark. Participants will learn how to use Spark NLP pipelines for preprocessing, named entity recognition (NER), sentiment analysis, and custom model development. The course emphasizes hands-on learning and scalable NLP solutions.

Introduction to Spark NLP and Pretrained Pipelines

Introduction to Spark NLP

• What is Spark NLP and why it’s used
• Spark NLP vs other NLP libraries (SpaCy, NLTK, etc.)
• Real-world applications and industry use cases
• Setting up Spark NLP (local, Colab, or Databricks)

Spark NLP Pipeline Architecture

• Understanding Spark NLP pipelines and annotators
• Key components: DocumentAssembler, Tokenizer, Normalizer, Lemmatizer
• Creating and chaining annotators in a pipeline
• Execution of pipelines using Spark DataFrames

Using Pretrained Pipelines and Models

• Introduction to pre-trained pipelines
• Named Entity Recognition (NER) with pre-trained models
• POS tagging, dependency parsing, and sentiment analysis
• Running pipelines on structured and unstructured text

Practical Exercises with Pretrained Pipelines

• Load and apply pipelines on sample datasets
• Extract entities, sentiment, and key phrases
• Save and load trained pipelines
• Output visualization and result interpretation

Custom Pipelines, Model Training, and Real-World Exercises

Trainable Annotators and Custom Pipelines

• Overview of trainable vs non-trainable annotators
• Building a pipeline from scratch
• Adding embeddings: WordEmbeddings, BERTEmbeddings
• Using multi-language support in pipelines

Training Custom NER Models

• Preparing data for NER model training
• Using NerDLApproach for deep learning-based NER
• Configuring training parameters
• Evaluating the custom model (precision, recall, F1-score)
• Saving and reusing custom models

Text Classification with Spark NLP

• Training a multi-class or binary classifier
• Using ClassifierDLApproach
• Input formats and labeling strategies
• Evaluating and exporting the classifier model

Real-World Exercises and Deployment

• Wrap-up and next steps (resources, career paths, certification)