MoralBERT: A Fine-Tuned Language Model for Capturing Moral Values in Social Discussions

• URL: http://arxiv.org/abs/2403.07678v2
• Date: Fri, 19 Jul 2024 15:27:35 GMT
• Title: MoralBERT: A Fine-Tuned Language Model for Capturing Moral Values in Social Discussions
• Authors: Vjosa Preniqi, Iacopo Ghinassi, Julia Ive, Charalampos Saitis, Kyriaki Kalimeri,
• Abstract summary: Moral values play a fundamental role in how we evaluate information, make decisions, and form judgements around important social issues. Recent advances in Natural Language Processing (NLP) show that moral values can be gauged in human-generated textual content. This paper introduces MoralBERT, a range of language representation models fine-tuned to capture moral sentiment in social discourse.
• Score: 4.747987317906765
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Moral values play a fundamental role in how we evaluate information, make decisions, and form judgements around important social issues. Controversial topics, including vaccination, abortion, racism, and sexual orientation, often elicit opinions and attitudes that are not solely based on evidence but rather reflect moral worldviews. Recent advances in Natural Language Processing (NLP) show that moral values can be gauged in human-generated textual content. Building on the Moral Foundations Theory (MFT), this paper introduces MoralBERT, a range of language representation models fine-tuned to capture moral sentiment in social discourse. We describe a framework for both aggregated and domain-adversarial training on multiple heterogeneous MFT human-annotated datasets sourced from Twitter (now X), Reddit, and Facebook that broaden textual content diversity in terms of social media audience interests, content presentation and style, and spreading patterns. We show that the proposed framework achieves an average F1 score that is between 11% and 32% higher than lexicon-based approaches, Word2Vec embeddings, and zero-shot classification with large language models such as GPT-4 for in-domain inference. Domain-adversarial training yields better out-of domain predictions than aggregate training while achieving comparable performance to zero-shot learning. Our approach contributes to annotation-free and effective morality learning, and provides useful insights towards a more comprehensive understanding of moral narratives in controversial social debates using NLP.

Related papers

• Automatic Detection of Moral Values in Music Lyrics [4.747987317906765]
  Moral values play a fundamental role in how we evaluate information, make decisions, and form judgements around important social issues. We tasked a set of transformer-based language models (BERT) fine-tuned on 2,721 synthetic lyrics to detect moral values in 200 real music lyrics annotated by two experts. We evaluate their predictive capabilities against a series of baselines including out-of-domain (BERT fine-tuned on MFT-annotated social media texts) and zero-shot (GPT-4) classification. The proposed models yielded the best accuracy across experiments, with an average F1 weighted score of 0.8.
  arXiv  Detail & Related papers  (2024-07-26T14:49:21Z)
• What Makes it Ok to Set a Fire? Iterative Self-distillation of Contexts and Rationales for Disambiguating Defeasible Social and Moral Situations [48.686872351114964]
  Moral or ethical judgments rely heavily on the specific contexts in which they occur. We introduce defeasible moral reasoning: a task to provide grounded contexts that make an action more or less morally acceptable. We distill a high-quality dataset of 1.2M entries of contextualizations and rationales for 115K defeasible moral actions.
  arXiv  Detail & Related papers  (2023-10-24T00:51:29Z)
• Value Kaleidoscope: Engaging AI with Pluralistic Human Values, Rights, and Duties [68.66719970507273]
  Value pluralism is the view that multiple correct values may be held in tension with one another. As statistical learners, AI systems fit to averages by default, washing out potentially irreducible value conflicts. We introduce ValuePrism, a large-scale dataset of 218k values, rights, and duties connected to 31k human-written situations.
  arXiv  Detail & Related papers  (2023-09-02T01:24:59Z)
• Rethinking Machine Ethics -- Can LLMs Perform Moral Reasoning through the Lens of Moral Theories? [78.3738172874685]
  Making moral judgments is an essential step toward developing ethical AI systems. Prevalent approaches are mostly implemented in a bottom-up manner, which uses a large set of annotated data to train models based on crowd-sourced opinions about morality. This work proposes a flexible top-down framework to steer (Large) Language Models (LMs) to perform moral reasoning with well-established moral theories from interdisciplinary research.
  arXiv  Detail & Related papers  (2023-08-29T15:57:32Z)
• Towards Few-Shot Identification of Morality Frames using In-Context Learning [24.29993132301275]
  We study few-shot identification of a psycho-linguistic concept, Morality Frames, using Large Language Models (LLMs) Morality frames are a representation framework that provides a holistic view of the moral sentiment expressed in text. We propose prompting-based approaches using pretrained Large Language Models for identification of morality frames, relying on few-shot exemplars.
  arXiv  Detail & Related papers  (2023-02-03T23:26:59Z)
• ClarifyDelphi: Reinforced Clarification Questions with Defeasibility Rewards for Social and Moral Situations [81.70195684646681]
  We present ClarifyDelphi, an interactive system that learns to ask clarification questions. We posit that questions whose potential answers lead to diverging moral judgments are the most informative. Our work is ultimately inspired by studies in cognitive science that have investigated the flexibility in moral cognition.
  arXiv  Detail & Related papers  (2022-12-20T16:33:09Z)
• The Moral Foundations Reddit Corpus [3.0320832388397827]
  Moral framing and sentiment can affect a variety of online and offline behaviors. We present the Moral Foundations Reddit Corpus, a collection of 16,123 Reddit comments curated from 12 distinct subreddits.
  arXiv  Detail & Related papers  (2022-08-10T20:08:10Z)
• Learning to Adapt Domain Shifts of Moral Values via Instance Weighting [74.94940334628632]
  Classifying moral values in user-generated text from social media is critical to understanding community cultures. Moral values and language usage can change across the social movements. We propose a neural adaptation framework via instance weighting to improve cross-domain classification tasks.
  arXiv  Detail & Related papers  (2022-04-15T18:15:41Z)
• Identifying Morality Frames in Political Tweets using Relational Learning [27.047907641503762]
  Moral sentiment is motivated by its targets, which can correspond to individuals or collective entities. We introduce morality frames, a representation framework for organizing moral attitudes directed at different entities. We propose a relational learning model to predict moral attitudes towards entities and moral foundations jointly.
  arXiv  Detail & Related papers  (2021-09-09T19:48:57Z)
• Aligning AI With Shared Human Values [85.2824609130584]
  We introduce the ETHICS dataset, a new benchmark that spans concepts in justice, well-being, duties, virtues, and commonsense morality. We find that current language models have a promising but incomplete ability to predict basic human ethical judgements. Our work shows that progress can be made on machine ethics today, and it provides a steppingstone toward AI that is aligned with human values.
  arXiv  Detail & Related papers  (2020-08-05T17:59:16Z)
• Text-based inference of moral sentiment change [11.188112005462536]
  We present a text-based framework for investigating moral sentiment change of the public via longitudinal corpora. We build our methodology by exploring moral biases learned from diachronic word embeddings. Our work offers opportunities for applying natural language processing toward characterizing moral sentiment change in society.
  arXiv  Detail & Related papers  (2020-01-20T18:52:45Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.