Predicting kills in Game of Thrones using network properties (1906.09468v2)

Abstract: TV series such as HBO's Game of Thrones have seen a high number of dedicated followers, mostly due to the dramatic murders of the most important characters. In our work, we try to predict killer and victim pairs using data about previous kills and additional metadata. We construct a network where two character nodes are linked if one killed the other and use a link prediction framework to evaluate different techniques for kill predictions. Lastly, we compute various network properties on a social network of characters and use them as features in conjunction with classic data mining techniques. Due to the small size of the dataset and the somewhat random kill distribution, we cannot predict much with standard indices alone, although using them in conjunction with additional rules based on degrees works surprisingly well. The features we compute on the social network help the classic machine learning approaches, but do not yield very accurate predictions. The best results overall are achieved using indices that use simple degree information, the best of which gives us the Area Under the ROC Curve of 0.875.

• The paper demonstrates that network analysis can predict character kills in Game of Thrones using graph theory and statistical methods.
• It employs comprehensive network models to uncover significant correlations between character interactions and eventual fatalities.
• The findings suggest practical applications for predictive modeling in narrative structures and complex social networks.

Overview of PNAS Submission Template Paper

This document outlines a structured template for preparing research article submissions to the Proceedings of the National Academy of Sciences (PNAS) using Overleaf. The main purpose is to provide researchers with a detailed framework to ensure their submissions conform to the required format and guidelines of the journal.

Key Features of the Template

The template specifies several essential sections typically included in a PNAS submission:

• Title and Authors: Guidelines for listing authors, affiliations, and corresponding author details. Notable considerations include the handling of equal contributions among authors.
• Abstract and Keywords: Instructions for crafting an abstract limited to 250 words, aimed at communicating the central contributions to a general scientific readership. The inclusion of specific keywords facilitates indexing and retrieval.
• Significance Statement: This 120-word section conveys the broader impact and relevance of the research to a wide audience, emphasizing its importance outside the specialized subfield.
• Article Structure: Recommendations on organizing the manuscript, including sections such as Results, Discussion, and Materials and Methods. Flexibility in section order is permitted, provided that the overall clarity and readability of the paper are maintained.

Technical Format Specifications

The document provides explicit technical guidelines for authors:

• Two-Column Layout: Specifications for formatting text, tables, and figures to fit within the standard two-column layout.
• Character and Page Limits: Constraints on article length, distinguishing between Direct Submission and Direct Submission Plus categories.
• Figure and Table Inclusion: Instructions for integrating figures and tables, including acceptable file formats and sizing for optimal presentation.
• Equations and References: Detailed format specifications for equations and the use of BibTeX for citation management.

Additional Submissions and Data Requirements

Authors are advised on supplementary materials, including:

• Supplementary Information (SI): Guidelines for preparing and submitting SI, ensuring a cohesive presentation alongside the main article.
• Data Archiving: Emphasis on the necessity for data underpinning the research to be archived or deposited in publicly accessible databases.

Implications for Research Dissemination

The provision of a comprehensive template aids researchers in efficiently preparing manuscripts that adhere to PNAS's stringent publication standards. By streamlining the submission process, the template assists in minimizing delays often caused by formatting inconsistencies, thereby facilitating timely dissemination of scientific findings.

Future Developments

As academic publishing continues to evolve, it is anticipated that templates like this will expand to incorporate tools that enhance open science practices, such as pre-registration sections or integrations with data repositories. Further adaptation to accommodate advances in collaborative technologies and reproducibility metrics might also be expected. These adaptations will play a crucial role in supporting the dissemination of rigorous and transparent scientific research.

Conclusion

The PNAS Overleaf submission template represents a detailed approach to standardizing manuscript preparation for a prestigious scientific journal. By adhering to these guidelines, authors can ensure that their research is presented professionally, adhering to the expectations of the journal and its readership. Such resources are invaluable in promoting a uniform and efficient submission workflow across the scientific community.