Kewords: Natural Language Processing - Applied NLP Thinking - Network Analysis - Sentiment Analysis - Topic Modeling - Media Studies - Harry Potter - Learning Reflection
This paper describes a first approach to the topic of NLP, more precisely Applied NLP. With the help of the software products of the company explosion (spaCy and Prodigy), the basic features of NLP are to be comprehended and applied to an exemplary project topic.
For this purpose, data from the Harry Potter saga in the form of its film scripts was used in order to achieve initial analysis results in the question area of network analysis with the help of sentiment analysis and topic modeling. The project focuses on the learning experience by documenting and reflecting on the project and its implementations as well as the visualization of the analysis results. Thus, the implementation of the NLP question is based on rather rudimentary Python knowledge and the desire to gain first basic knowledge in the application and understanding of NLP software. For this purpose, first an outline of theoretical basics is recorded, in a further step the implementation is documented and finally the project experience is reflected.
The corresponding repository contains the data (hp_data), a folder with Python and Markdown basics (project_basics), a notebook with code and one with documentation, as well as the visualizations in separate folders (hp_sentimentDistribution, hp_network, hp_wordclouds), the database (hp-dataset-with-relations.sqlite), a textfile including the dialogtext (all_parts.txt) and this Readme (readme.txt).
@authors: Teresa Wendel, Merle-Sophie Thoma (January 31, 2022) Weiternutzung als OER ausdrücklich erlaubt: Dieses Werk und dessen Inhalte sind - sofern nicht anders angegeben - lizenziert unter CC BY 4.0. Nennung bitte wie folgt: "Harry Potter and a network analysis" von Merle-Sophie Thoma und Teresa Wendel, Lizenz: CC BY 4.0.