Towards an Event Based Plot Model (JCLS 2022)

Overview and ressources

This repository provides annotation data, analysis results and visualization results for our contribution to JCLS2022. Specifically, there are the following resources:

• ./event_annotations_with_summary_data/: event annotations for our four manually annotated texts icluding the freqeuncy of summaries that mention a specific event
• narrpy a python package used for the provided visualizations
• ./summary_annotations/: the manual summary annotations
• ./summary_optimization_results/: a json file which provides the data our optimization results are based on
• jcls2022.ipynb: a Jupyter Notebook that provides the visulizations of our paper in an interactive version

Prerequisites

narrpyand the Jupyter Notebook have the following requirements:

• numpy==1.21.0
• seaborn==0.11.2
• matplotlib==3.4.3
• matplotlib-inline==0.1.3
• pandas==1.3.4
• plotly==5.3.1