Modules of the Workshop : Day-1

Module – 1 : Introduction to NLP

Concepts Covered :

• Basics of natural language processing (NLP) and understanding text data.

• Preprocessing techniques: tokenization, stemming, lemmatization, word similarity.

Learning Outcomes :

Ability to process and clean text data using fundamental preprocessing techniques.

Module – 2 : Word Embedding Models

Concepts Covered :

• Hyperspace Analogue to Language, COALS, GloVe, Word2Vec. 
• Convert words to vectors, similar words, word vector operations.

Learning Outcomes :

Understand and apply word embeddings to represent and analyze textual data.

Module – 3 : N-gram Language Models

Concepts Covered :

• Introduction to language models, context in modeling, chain rule, Markov assumption. 
• n-gram language models, perplexity, and quality evaluation.

Learning Outcomes :

Build and evaluate probabilistic language models.

Module – 4 : Neural Language Models

Concepts Covered :

• Disadvantages of probabilistic models and curse of dimensionality. 
• Neural models for word embedding and CBOW.

Learning Outcomes :

Understand neural network approaches to build language models.

Modules of the Workshop : Day - 2

Module – 1 : Recurrent Neural Networks (RNNs)

Concepts Covered :

• Challenges in traditional neural networks for NLP. 
• Introduction to RNNs, variable-length sequences, training RNNs for language modeling.

Learning Outcomes :

Develop language models for handling variable-length sequences using RNNs.

Module – 2 : Deep Contextualized Models

Concepts Covered :

• Deep contextualized word representation with ELMo.
• Importance of contextualized models.

Learning Outcomes :

Learn the importance and application of contextualized language models.

Module – 3 : Large Language Models

Concepts Covered :

Motivation behind Transformers, attention mechanism, self-attention, multi-head attention.

Learning Outcomes :

Understand the architecture and significance of Transformers in language modeling.

Module – 4 : BERT and GPT

Concepts Covered :

• BERT and GPT models for language modeling. 
• Overview of state-of-the-art advancements.

Learning Outcomes :

Comprehend state-of-the-art language models and their applications.

Module – 5 : Bias, Ethics, and Advances in NLP

Concepts Covered :

• Ethical considerations in NLP, including bias in language models. 
• Recent advances in language models and their applications.

Learning Outcomes :

Understand ethical challenges in NLP and explore the latest advancements in language modeling.