Web resources in linked open data (LOD) are comprehensible to humans through literal textual values attached to them, such as labels, notes, or comments. Word choices in literals may not always be neutral. When culturally stereotyping terminology is used in literals, they may appear as offensive to users in interfaces and propagate stereotypes to algorithms trained on them. We study how frequently and in which literals contentious terms about people and cultures occur in LOD and whether there are attempts to mark the usage of such terms. For our analysis, we reuse English and Dutch terms from a knowledge graph that provides opinions of experts from the cultural heritage domain about terms' contentiousness. We inspect occurrences of these terms in four widely used datasets: Wikidata, The Getty Art & Architecture Thesaurus, Princeton WordNet, and Open Dutch WordNet. Some terms are ambiguous and contentious only in particular senses. Applying word sense disambiguation, we generate a set of literals relevant to our analysis. We found that contentious terms frequently appear in descriptive and labelling literals, such as preferred labels that are usually displayed in interfaces and used for indexing. In some cases, LOD contributors mark contentious terms with words and phrases in literals (implicit markers) or properties linked to resources (explicit markers). However, such marking is rare and non-consistent in all datasets. Our quantitative and qualitative insights could be helpful in developing more systematic approaches to address the propagation of stereotypes via LOD.

Recent initiatives by cultural heritage institutions in addressing outdated and offensive language used in their collections demonstrate the need for further understanding into when terms are problematic or contentious. This paper presents an annotated dataset of 2,715 unique samples of terms in context, drawn from a historical newspaper archive, collating 21,800 annotations of contentiousness from expert and crowd workers. We describe the contents of the corpus by analysing inter-rater agreement and differences between experts and crowd workers. In addition, we demonstrate the potential of the corpus for automated detection of contentiousness. We show that a simple classifier applied to the embedding representation of a target word provides a better than baseline performance in predicting contentiousness. We find that the term itself and the context play a role in whether a term is considered contentious.