Paper: Social Emotion Mining Techniques for Facebook Posts Reaction Prediction (accepted at ICAART 2018)

This dataset was created for a research project at the Department of Data Science and Knowledge Engineering (Maastricht University). The dataset contains facebook posts, their correlating comments, an emotion lexicon and labled sentences.

The database used in this project is MongoDB. The files in this repository are dump-files created by MongoDB. So, one can unzip the files, start MongoDB on your machine and use the following command to import the files in your MongoDB:

mongorestore -d <name_of_the_database> <your_path_to_the_github_files>

We also provide Python 3 scripts that can be used to work with the data. One needs to install pymongo to use these scripts.

DataStorage is a base class (abstract class) for accessing/reading/writing to MongoDB.

MongodbStorage inherits from DataStorage and contains the implemented database access (default database name is "research_project", the user should use <name_of_the_database> as specified in the import command). There are various methods to read and write information to the database tables.

Furthermore, Post, Comment and Emotion are three data classes that can hold information of the corresponding database tables.