The extended Moral Foundations Dictionary (eMFD): Development and applications of a crowd-sourced approach to extracting moral intuitions from text

@article{Hopp2020TheEM,
  title={The extended Moral Foundations Dictionary (eMFD): Development and applications of a crowd-sourced approach to extracting moral intuitions from text},
  author={Frederic R. Hopp and Jacob T. Fisher and Devin Cornell and Richard Huskey and Ren{\'e} Weber},
  journal={Behavior Research Methods},
  year={2020},
  volume={53},
  pages={232-246}
}
Moral intuitions are a central motivator in human behavior. Recent work highlights the importance of moral intuitions for understanding a wide range of issues ranging from online radicalization to vaccine hesitancy. Extracting and analyzing moral content in messages, narratives, and other forms of public discourse is a critical step toward understanding how the psychological influence of moral judgments unfolds at a global scale. Extant approaches for extracting moral content are limited in… 

Figures from this paper

The Moral Foundations Reddit Corpus

The Moral foundations Reddit Corpus is presented, a collection of 16,123 English Reddit comments that have been curated from 12 distinct subreddits, hand-annotated by at least three trained annotators for 8 categories of moral sentiment based on the updated Moral Foundations Theory (MFT) framework.

Uncovering Values: Detecting Latent Moral Content from Natural Language with Explainable and Non-Trained Methods

This work develops and compares two different approaches: a frame-based symbolic value detector based on knowledge graphs and a zero-shot machine learning model fine-tuned on a task of Natural Language Inference (NLI) and atask of emotion detection.

Do Bots Have Moral Judgement? The Difference Between Bots and Humans in Moral Rhetoric

This work analyzes the evolution of latent moral loadings over time and across different sub-narratives on human and bot-generated tweets from April 1st, 2018 to April 30th, 2019 to present the significant differences between the strength and patterns of moral rhetoric for human andBot-generated content on Twitter.

LibertyMFD: A Lexicon to Assess the Moral Foundation of Liberty.

Quantifying the moral narratives expressed in the user-generated text, news, or public discourses is fundamental for understanding individuals’ concerns and viewpoints and preventing violent protests

Incorporating the Measurement of Moral Foundations Theory into Analyzing Stances on Controversial Topics

The correlation analysis of stance and morality revealed notable associations between stances on social issues and various types of morality, such as care, fairness, and loyalty, hence demonstrating that there are certain morality types that are more attributed to stance classification than others.

Cross-Domain Classification of Moral Values

It is shown that a value classifier can generalize and transfer knowledge to novel domains, but it can introduce catastrophic forgetting, and the typical classification errors in cross-domain value classification are highlighted.

Moral Mimicry: Large Language Models Produce Moral Rationalizations Tailored to Political Identity

Large Language Models (LLMs) have recently demonstrated impressive capability in generating fluent text. LLMs have also shown an alarming tendency to reproduce social biases, for example stereotypical

Noise Audits Improve Moral Foundation Classification

Two metrics to audit the noise of annotations are proposed and experiments show that removing noisy annotations based on the proposed metrics improves classification performance.

Learning to Adapt Domain Shifts of Moral Values via Instance Weighting

This is the first study to quantify impacts of moral shifts, propose adaptive framework to model the shifts, and conduct a case study to model COVID-19 vaccine-related behaviors from moral values.

A Graph-Learning Approach for Detecting Moral Conflict in Movie Scripts

An approach combining tools from social network analysis and natural language processing with recent theoretical advancements in the Model of Intuitive Morality and Exemplars is presented, showing that scenes containing moral conflict between central characters can be identified using changes in connectivity patterns between network modules.

References

SHOWING 1-10 OF 80 REFERENCES

Extracting Latent Moral Information from Text Narratives: Relevance, Challenges, and Solutions

ABSTRACT Moral Foundations Theory (MFT) and the Model of Intuitive Morality and Exemplars (MIME) contend that moral judgments are built on a universal set of basic moral intuitions. A large body of

Enhancing the Measurement of Social Effects by Capturing Morality

This work empirically evaluates the usage of a morality lexicon that was expanded via a quality controlled, human in the loop process and finds that the enhancement of the original lexicon led to measurable improvements in prediction accuracy for the selected NLP tasks.

Morality Between the Lines : Detecting Moral Sentiment In Text

This paper uses shortpost social media to compare the accuracy of text analysis methods for detecting moral rhetoric and longer form political speeches to explore detecting shifts in that rhetoric over time, and shows that combining the MFD with distributed representations allows for a cleaner signal when detectingmoral rhetoric, particularly with shortform text.

Moral Foundations Twitter Corpus: A Collection of 35k Tweets Annotated for Moral Sentiment

The Moral Foundations Twitter Corpus is introduced, a collection of 35,108 tweets that have been curated from seven distinct domains of discourse and hand annotated by at least three trained annotators for 10 categories of moral sentiment.

A Graph-Learning Approach for Detecting Moral Conflict in Movie Scripts

An approach combining tools from social network analysis and natural language processing with recent theoretical advancements in the Model of Intuitive Morality and Exemplars is presented, showing that scenes containing moral conflict between central characters can be identified using changes in connectivity patterns between network modules.

Sentiment analysis of political communication: combining a dictionary approach with crowdcoding

A procedure for collecting fine-grained sentiment scores through crowdcoding to build a negative sentiment dictionary in a language and for a domain of choice is presented and the results show that the crowdbased dictionary provides efficient and valid measurement of sentiment.

Measuring Moral Rhetoric in Text

A computational text analysis technique for measuring the moral loading of concepts as they are used in a corpus, using latent semantic analysis to compute the semantic similarity between concepts and moral keywords taken from the “Moral foundation Dictionary”.

The MAD Model of Moral Contagion: The Role of Motivation, Attention, and Design in the Spread of Moralized Content Online

Evidence that expressions of moral emotion play an important role in the spread of moralized content and a psychological model called the motivation, attention, and design (MAD) model is proposed to explain moral contagion.

The Morality of May 2, 2011: A Content Analysis of U.S. Headlines Regarding the Death of Osama bin Laden

Following the death of Osama bin Laden in the late hours of May 1, 2011, many print newspapers throughout the United States and the world ran front-page coverage of his death the following day.
...