Natural Language Processing

The article is designed for readers with some conceptual grasp of computer science who wish to develop their understanding of Artificial Intelligence and more specifically NATURAL LANGUAGE PROCESSING (NLP).

CONCEPTS : Computer science, Machine learning, Neural Networks, Natural language Understanding.

Elaine Rich in her text book artificial intelligence (1988) outlines the problem of natural language understanding even in the more simple context of written language, describing it as extremely difficult. lt;/p>

To quote ` In order to understand sentences about about a topic, it is necessary to know not only a lot about a language itself (vocabulary, grammar) but a great deal about the topic so that the unstated assumptions can be recognised.`

The initial funding for the development of this article came from an enumeration for research at Kings College London during the cold winter of Nov 2024. The NHS hospital site was based near Denmark Hill London, not far from Southbank University, initiators of my status in Artificial Intelligence. Another lucky break came in early autumn 2025 with finance (love2shop tokens possibly equivalent to representative money) for focus group participation in PhD research at UCL London, the scientific abstract mentioned future developments will include training an NLP algorithm to generate lay summaries from medical documentation.

Alan Turing`s published academic paper `Computing Machinery and Intelligence, the article proposed what is know THE TURING TEST as a criterion of intelligence, the test now involves the automated interpretation and generation of NATURAL LANGUAGE, it is so respectable it has been referred to on BBC TV children`s programme BLUE PETER.

WIKIPEDIA provides a comprehensive chronology of evens defining developments in Natural Language Processing.

Initially RULE BASE SYSTEMS (symbolic approach - ie hand coding rules) provide the automated translation interface, these became increasingly obsolete with with he advance of STATISTICAL or NEURAL NETWORKS, these are more robust to unfamiliar and erroneous input (misspelled or accidentally omitted words). the larger PROBABILISTIC models become more accurate as they perform unlike RULE BASED SYSTEMS.