This library provides the code for a flexible calculation of moral foundation scores from textual input data. We have developed this framework inspired by FrameAxis paper.
Using this framework we have studied two applications:
- Moral Framing and Ideological Bias of News, 2020
- Mapping Moral Valence of Tweets Following the Killing of George Floyd, 2021
If you are using this code please consider citing our papers and giving this repository a star.
FrameAxis is a flexible framework and all it needs is sets of antonym words. There are several moral foundation dictionaries that provide vice and virtues of moral categories. In this code we are supporting:
- MFD: The original moral foundations dictionary proposed by Graham et al., 2009
- MFD2.0: A moral foundations dictionary proposed by Frimer et al., 2017
- eMFD: An extended moral foundations dictionary extracted from human-annotated corpora by Hopp et al., 2021.
- customized: If you don't want to stick to these dictionaries; maybe you want to curate your own dictionary in your language or go out of the scope of moral foundations you can choose the dictionary argument as "--dict_type custom" and provide your file in moral_foundations_dictionary folder. Your dictionary file must be named as "custom.csv" and must have three columns named as "word", "category", and "sentiment". Same as the MFD_original.csv which is already provided.
-
[--input_file]: Path to the dataset .csv file containing input text documents in a column.
-
[--docs_colname]: The name of the column in the input file that contains the texts to calculate the MF scores on.
-
[--dict_type]: Dictionary for calculating FrameAxis Scores. Possible values are: emfd, mfd, mfd2, and customized.
-
[--word_embedding_model]: Path to the word embedding model used to map words to a vector space. If not specified a default w2v model will be used.
-
[--output_file]: The path for saving the MF scored output CSV file. The output file contains columns for MF scores concatenated to the original dataset.
IN_FILE="path_to_your_input.csv"
COLNAME="colname_to_calc_scores_on"
DICT="mfd"
OUT_FILE="path_to_save_results.csv"
python main.py --input_file $IN_FILE --docs_colname $COLNAME --dict_type $DICT --output_file $OUT_FILE