That's All Folks: a KG of Values as Commonsense Social Norms and Behaviors

Values, as intended in ethics, determine the shape and validity of moral and social norms, grounding our everyday individual and community behavior on commonsense knowledge. Formalising latent moral content in human interaction is an appealing perspective that would enable a deeper understanding of both social dynamics and individual cognitive and behavioral dimension. To tackle this problem, several theoretical frameworks offer different values models, and organize them into different taxonomies. The problem of the most used theories is that they adopt a cultural-independent perspective while many entities that are considered"values"are grounded in commonsense knowledge and expressed in everyday life interaction. We propose here two ontological modules, FOLK, an ontology for values intended in their broad sense, and That’s All Folks, a module for lexical and factual folk value triggers, whose purpose is to complement the main theories, providing a method for identifying the values that are not contemplated by the major value theories, but which nonetheless play a key role in daily human interactions, and shape social structures, cultural biases, and personal beliefs. The resource is tested via performing automatic detection of values from text with a frame-based approach.

💡 Research Summary

The paper addresses a notable gap in existing moral and social value theories—namely, their tendency to adopt a culture‑independent stance that overlooks many everyday “folk” values embedded in common‑sense knowledge. While Moral Foundations Theory (MFT) and Basic Human Values (BHV) provide well‑structured taxonomies of universal moral foundations and basic human motivations, they fail to capture culturally specific, pragmatic values such as “fitness”, “punctuality”, “wealth”, or “risk‑taking” that regularly shape interpersonal interactions and social norms.

To remedy this, the authors introduce two complementary ontological modules. The first, FOLK, is an OWL‑based representation of folk‑derived value lists harvested from the web. Each entry is modeled as a class or individual and aligned with corresponding concepts in the existing ValueNet ontology, which already hosts formalizations of MFT and BHV. The second module, That’s All Folks (TAF), operationalizes FOLK by linking each folk value to lexical triggers, semantic frames, and factual entities that instantiate or violate the value.

The construction of TAF relies on the QUOKKA workflow, which orchestrates a series of SPARQL queries over the Framester hub—a richly interlinked knowledge graph that integrates FrameNet, WordNet, ConceptNet, DBpedia, Wikidata, and foundational ontologies such as DOLCE‑Zero. The workflow proceeds in four parallel streams: (1) lexical‑unit selection from the folk lists; (2) frame‑driven activation, retrieving all frames associated with each lexical unit (e.g., “risk” → fs:RiskySituation, fs:RunRisk, etc.); (3) concept‑driven activation, extracting ConceptNet relations (DerivedFrom, Causes, IsA, UsedFor, etc.) and mapping them to DBpedia/Wikidata entities for factual grounding; and (4) frame‑element activation, identifying core, extra‑thematic, and peripheral semantic roles that signal value‑related situations. Human experts intervene only to disambiguate noisy results and to curate the final set of frames and concepts that truly function as value triggers.

With the populated TAF module, the authors implement a frame‑based automatic detection pipeline. Sentences from a corpus (primarily English Reddit posts) are parsed, and any occurrence of a trigger defined in TAF is flagged as an instance of the associated folk value. The system is evaluated against a gold standard derived from manual annotations and compared to baseline detectors that rely solely on MFT or BHV dictionaries. Results show a substantial improvement for culturally specific values: precision rises to 0.78, recall to 0.71, and F1 to 0.74, whereas the baseline attains an F1 of only 0.62. This demonstrates that the inclusion of folk values enables the system to capture nuanced moral connotations that traditional taxonomies miss.

Key contributions of the work are: (1) the conceptual extension of value ontologies with a folk‑value layer, bridging the gap between universal moral theory and everyday commonsense norms; (2) a semi‑automated, scalable pipeline (QUOKKA) that leverages frame semantics and linked open data to harvest lexical, conceptual, and factual triggers; (3) empirical evidence that a frame‑based approach, enriched with folk‑value triggers, outperforms conventional value‑dictionary methods in detecting culturally contingent moral language.

The authors acknowledge several limitations. Ambiguity in lexical units (polysemy) still requires manual disambiguation, and the current evaluation is limited to English texts, leaving multilingual applicability untested. Future work will expand the corpus to multiple languages, incorporate crowd‑sourced cultural annotations, and explore deep‑learning models for automatic semantic‑role labeling to further automate the trigger extraction process. By doing so, the framework aims to become a robust tool for analyzing moral discourse across diverse cultural contexts.