Resources and scripts for complexity & subjectivity project Katharina Ehret and Maite Taboada
This repository contains the following resources:
- Subjectivity. This folder contains the lists of argumentation and subjectivity markers used for measuring the concept subjectivity in Ehret & Taboada (2021). It contains two subfolders
other_features: stance adverbials.csv, connectives.csv and modals.csv
socal_features: socal_invariant, socal_variant
other_features contains lists of stance adverbials, connectives and modals which are "invariant features", i.e. they cannot take different forms.
socal_features contains lists of positive and negative adverbs, which classify as "invariant features" as well as negative and positive adjectives, nouns and verbs, which classify as "variant features", i.e. they can take different forms such as plural, comparative or third person singular. All positive and negative features were taken from SOCAL (REF) and have a valency of |4| and |5|.
- Scripts. This folder contains the scripts for data generation, clean-up and the retrieval of the subjectivity markers.
sub_marker.py: This script retrieves the raw text frequencies of the above listed argumentation and subjectivity markers.
Reference:
- Ehret, K. and M. Taboada (2021) The interplay of complexity and subjectivity in opinionated discourse. Discourse Studies 23(2): 141-165.