Textual Analysis with Natural Language Processing

Abstract

Since the development of graph theory 300 years ago, networks have been used to represent a multitude of systems from transport infrastructures to groups of people. A more recent application has been that of mythological texts. These networks, constructed of characters and their interactions, quantitatively tell of the narrative and allow it to be analysed using different graph theory measures. This analysis facilitates the comparison of different narratives and is employed within this work to evaluate multiple texts.In current times, artificial intelligence has bled into the many facets of academia, industry and general everyday life. This work utilises an area of artificial intelligence known as natural language processing to automate the extraction of social network data. The methodology within this study is developed with an Icelandic narrative and later applied to texts of other backgrounds to judge its capabilities.A further application of natural language processing to industry that is discussed in this study uses topic modelling to construct interactive networks of a company’s database. Instead of linking people by their interactions, this project links papers and authors via their topics and collaborations respectively and demonstrates other possible applications of natural language processing and networks.

Date of Award	May 2024
Original language	English
Awarding Institution	Coventry University
Supervisor	Ralph Kenna (Supervisor)