Beyond Topicality: A Conceptual Analysis of Societal Relevance and Its Application to Search Results and AI Responses

Collection of recent IR/ML papers examining (1) "societal relevance" as a framework for surfacing or suppressing content harmful to social interests (misinformation, discrimination) beyond topical relevance; (2) evolution of recommender systems from raw IDs to "semantic IDs" and a proposed semantic-planning paradigm; (3) multiple studies on automated content labelling (thematic indexing, keyword extraction) and associated stewardship/ethics trade-offs for crowdsourced collections; (4) explainability advances (Rashomon Explanation set and RashomonLLM) that couple explanation with prediction to,