0%
Preparing your experience
English 日本語
0 Start Learning

Your Bag

E-learning materials

Natural Language Processing (NLP)

Process and understand human language with AI, from tokenisation and embeddings to classification, extraction, and summarisation pipelines.

Material Overview
Natural Language Processing (NLP)

Learn how machines work with text. You will start with tokenisation, stemming, and classic representations, move on to embeddings and transformer models, and build practical pipelines for sentiment analysis, topic classification, named entity extraction, and summarisation. This E-Learning Material covers multilingual text, evaluation metrics, and the cases where a small fine tuned model beats a large general purpose one on both cost and accuracy.

Curriculum

Learning Path

Each level is enrolled separately. Work through them in order, or start at the level that matches your experience.

  1. Level 01

    Beginner

    Purpose

    Learn how text becomes something a model can work with, and build a first classifier over it.

    Outcome

    You can build a working text classifier and explain what it is keying on.

    What You'll Learn
    • Clean and tokenise text and understand what each step discards
    • Represent documents with counts and term weighting
    • Train a simple text classifier and inspect which words drove the decision

    275 CREDS

  2. Level 02

    Intermediate

    Purpose

    Use embeddings and pretrained models to capture meaning rather than matching words.

    Outcome

    You can apply pretrained language models to a specific task with the data you actually have.

    What You'll Learn
    • Generate embeddings and measure similarity between texts
    • Fine tune a pretrained transformer for classification on your own labels
    • Compare a small specialised model with a large general one on cost and accuracy

    476 CREDS

  3. Level 03

    Advanced

    Purpose

    Build extraction and summarisation pipelines that produce structured output from unstructured text.

    Outcome

    You can turn free text into structured records that downstream systems can use.

    What You'll Learn
    • Extract entities, relations, and attributes into a defined schema
    • Summarise long documents while keeping facts and figures intact
    • Handle multilingual text and the failure modes that come with translation

    755 CREDS

  4. Level 04

    Expert

    Purpose

    Evaluate and operate language systems where errors are subtle and easy to miss.

    Outcome

    You can keep a language processing system accurate as the text it sees keeps changing.

    What You'll Learn
    • Build annotated test sets and measure quality beyond a single accuracy score
    • Detect data drift as vocabulary and topics change over time
    • Balance model size, latency, and cost for continuous processing

    1,099 CREDS