Natural Language Processing

What Natural Language Processing is

NLP covers tokenisation, parsing, named entity recognition, sentiment analysis, translation, summarisation and question answering. Large language models absorbed many of these tasks into a single general system.

How it works

Classical pipelines chained rule-based and statistical components. Modern practice fine-tunes or prompts a pretrained transformer, though lightweight classical methods remain useful for cheap, high-volume preprocessing.

Why it matters

Most enterprise data is unstructured text, so NLP is how that mass of documents, tickets and emails becomes queryable and actionable.

Common uses

• →Document classification and routing
• →Entity and relationship extraction
• →Translation and localisation
• →Sentiment and intent analysis