||Hill Center, Room 359
||Appalchian State University
||The Complex Networks of Tolkien, Rowling, and Other Great Story-tellers
Many successful books and book series describe long stories that include a multitude of characters. The evolution of a story creates a timeline of events where book characters interact with each other. This timeline can then be translated into a dynamically evolving complex network. The analysis of complex networks is a rapidly expanding field and is ideally suited to extract information from complex interactions. In this case, the comparative analysis of the character affiliation networks in different books can be used to detect possible patterns that are either characteristic of a genre, a time-period for the genre, etc. Similarities and differences in story-building of different authors within and across genres are of particular importance and may hint to different writing styles.
The goal of the project is to introduce one or two students to the fundamental concepts of complex network theory and guide the students through the whole process of acquiring the data, apply the learned concepts to the data analysis, and decide on the important findings that should be highlighted in a research paper. This network-based project should demonstrate the power of applying complex network theory to a variety of disciplines.
- Week 1:
- Arrived on Sunday afternoon and got settled in. After orientation, I met with my mentor and discussed the project for the first time. I read two papers on complex network theory, and I put together a presentation on the project to present on Friday. I also began work on a simple data mining program to start extracting social networks from bodies of text.
- Week 2:
- Met again with Dr. Gallos and Dr. Nina Fefferman to recap the presentation from Friday. Discussed the next steps to take to work on the project. Continued work on the data mining program and began work on network analysis software.
- Week 3:
- Met with Dr. Gallos to discuss progress and direction for project. Continued work on data mining software. Demonstrated software that could read in a network and find the degree distribution. Began work on an algorithm for finding the largest cluster within a network.
- Week 4:
- Continued work on data mining program. Am nearing completion of software for extracting social networks from bodies of text. Wrote algorithm for extracting the giant component from a network and began work on analyzing the giant component's degree distribution and clustering coefficient.
- Week 5:
- Finished rough work on data mining program. Began refining program to remove non-name nodes from extracted network.
- Week 6:
- Extracted network from the first Lord of the Rings book. Continued work on software to perform analysis. Began putting networks into Pajek for visualization.
- Week 7:
- Prepared presentation, extracted two more networks from the second and thirds Lord of the Rings books.