MoralStrength: Exploiting a moral lexicon and embedding similarity for moral foundations prediction

Abstract

Moral rhetoric plays a fundamental role in how we perceive and interpret the information we receive, greatly influencing our decision-making process. Especially when it comes to controversial social and political issues, our opinions and attitudes are hardly ever based on evidence alone. The Moral Foundations Dictionary (MFD) was developed to operationalize moral values in the text. In this study, we present MoralStrength, a lexicon of approximately 1,000 lemmas, obtained as an extension of the Moral Foundations Dictionary, based on Word-Net synsets. Moreover, for each lemma it provides with a crowdsourced numeric assessment of Moral Valence, indicating the strength with which a lemma is expressing the specific value. We evaluated the predictive potentials of this moral lexicon, defining three utilization approaches of increased complexity, ranging from lemmas' statistical properties to a deep learning approach of word embeddings based on semantic similarity. Logistic regression models trained on the features extracted from MoralStrength, significantly outperformed the current state-of-the-art, reaching an F1-score of 87.6% over the previous 62.4% (p-value< 0.01), and an average F1-Score of 86.25% over six different datasets. Such findings pave the way for further research, allowing for an in-depth understanding of moral narratives in text for a wide range of social issues.

Deciding between moral priorities and COVID-19 avoiding behaviors: A moral foundations vignette study

“…Morality plays an important role in regulating individuals' behaviors in social life (Araque et al, 2019;Ellemers et al, 2013;Ellemers & van den Bos, 2012) by public condemnation of immoral behaviors (Prosser et al, 2020). Similarly, during the pandemic, non-compliant behaviors such as not wearing mask or not keeping physical distance have been moralized at a community level (Francis & McNabb, 2020;Prosser et al, 2020).…”

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

“…Recent data-driven approaches have sought to ameliorate the limitations inherent in using small (< 500) word lists created manually by individual experts. These are: (a) expanding the number of words in the wordcount dictionary by identifying other words that are semantically similar to the manually selected "moral" words (Frimer et al, 2017;Rezapour et al, 2019); (b) abandoning simple wordcount-based scoring procedures and instead assessing the semantic similarity of entire documents to manually selected seed words (see e.g., Araque, Gatti, & Kalimeri, 2019;Garten et al, 2018;Sagi & Dehghani, 2017). Although these approaches are promising due to their computational efficiency and ability to score especially short texts (e.g., social media posts), their usefulness is less clear for determining which textual stimuli evoke an intuitive moral response among individuals.…”

Moral concerns are differentially observable in language

“…The majority of the studies at the intersection of moral psychology and language (e.g., Araque, Gatti, & Kalimeri, 2020;Mokhberian, Abeliuk, Cummings, & Lerman, 2020;Mooijman et al, 2018;Rezapour, Shah, & Diesner, 2019) rely on MFT as a guiding framework. MFT provides a predictive and pluralistic view of moral concerns that can facilitate an exploration into how such concerns are manifested in language.…”

“…To measure moral language, we rely on dictionaries of moral words, which provide a priori measures of explicit, word-level indicators of target constructs (Pennebaker, Francis, & Booth, 2001) that have been widely used in previous works to measure moral language (e.g., Araque et al, 2020;Dehghani et al, 2016). There are known limitations of this approach, however.…”