NLP Foundation for Generative AI

Singapore University of Social Sciences

NLP Foundation for Generative AI (ICT304)

Applications Open: To be confirmed

Applications Close: To be confirmed

Next Available Intake: To be confirmed

Course Types: To be confirmed

Language: English

Duration: 6 months

Fees: To be confirmed

Area of Interest: Science and Technology

Schemes: To be confirmed

Funding: To be confirmed

School/Department: School of Science and Technology

Bookmark

Synopsis

In this course, we will explore the foundational components of generative AI systems and learn how to solve specific tasks using these technologies. The integration of generative AI into the future of work is becoming increasingly prevalent. While generative AI is a relatively new field, it builds upon the established discipline of Natural Language Processing (NLP), historically considered a key aspect of artificial intelligence. NLP encompasses a variety of tasks, such as textual analysis— including word similarity, entity recognition, and sentence summarization—and applications like question answering and dialogue systems. Mastering these tasks requires a level of intelligence and serves as the groundwork for contemporary generative AI systems.

Level: 3
Credit Units: 5
Presentation Pattern: EVERY REGULAR SEMESTER

Topics

N-gram and the Markov model
Part of speech tagging using statistical method
Regression models for textual classification
Forward algorithm and decoding algorithms
Computing likelihood and the Viterbi algorithm
Word, sentence and tokenizers
Word semantics and embeddings
Grammar: statistical analysis vs neural models
Finding the right data: a case in machine translation
The concept of mapping: a case in Q&A
The new paradigm: named entity recognition, summarisation
The new paradigm: dialogue systems

Learning Outcome

Demonstrate the understanding of computational theories for understanding language.
Discuss the strengths and weaknesses of the various computational theories in NLP
Identify the different components needed to solve an NLP task
Illustrate an NLP system pipeline for solving textual problems.
Apply the various techniques of textual analysis to achieve a practical NLP system
Evaluate the current system and fine-tune the system to fulfill a different task.
Collect and process data for training an NLP model
Design the training of NLP model from scratch