Social Biases in Knowledge Representations of Wikidata separates Global North from Global South

Abstract: Knowledge Graphs have become increasingly popular due to their wide usage in various downstream applications, including information retrieval, chatbot development, LLM construction, and many others. Link prediction (LP) is a crucial downstream task for knowledge graphs, as it helps to address the problem of the incompleteness of the knowledge graphs. However, previous research has shown that knowledge graphs, often created in a (semi) automatic manner, are not free from social biases. These biases can have harmful effects on downstream applications, especially by leading to unfair behavior toward minority groups. To understand this issue in detail, we develop a framework -- AuditLP -- deploying fairness metrics to identify biased outcomes in LP, specifically how occupations are classified as either male or female-dominated based on gender as a sensitive attribute. We have experimented with the sensitive attribute of age and observed that occupations are categorized as young-biased, old-biased, and age-neutral. We conduct our experiments on a large number of knowledge triples that belong to 21 different geographies extracted from the open-sourced knowledge graph, Wikidata. Our study shows that the variance in the biased outcomes across geographies neatly mirrors the socio-economic and cultural division of the world, resulting in a transparent partition of the Global North from the Global South.

An In-Depth Analysis of ACM's Consolidated Article Template

The paper, entitled "The Name of the Title Is Hope," provides a comprehensive overview of the ACM consolidated article template introduced in 2017. The document serves as a guide for authors, highlighting the features and functionalities of the "acmart" document class, which is pivotal for preparing manuscripts across ACM's diverse range of publications.

Template Versatility and Functionality

The "acmart" document class is designed with versatility in mind, catering to varied types of submissions, including double-anonymous initial submissions, camera-ready journal articles, and conference proceedings. Three journal styles (acmsmall, acmlarge, and acmtog) and multiple conference styles (sigconf, sigchi, sigplan) provide flexibility in formatting, ensuring alignment with publication-specific requirements. This adaptability streamlines the submission process for authors who must navigate different submission types and presentation needs.

Prohibitions and Mandated Elements

A noteworthy aspect of the paper is the emphasis on maintaining integrity and consistency within the submission using strict formatting guidelines. Modifications to margins, typeface sizes, and spacing, among others, are expressly prohibited. The Libertine typeface family is mandated across submissions, ensuring uniformity and adherence to ACM's stylistic standards. Such constraints, while rigorous, are crucial in maintaining a professional and standardized presentation that facilitates review and publication.

Metadata and Rights Management

Critical to the digital dissemination and indexing of articles is the inclusion of metadata for accessibility and search optimization. Authors are guided through the necessary steps to incorporate ACM’s Computing Classification System (CCS) concepts and custom keywords, bolstering the discoverability of their work. Further, the paper advises on handling rights information, a prerequisite for final publication, which impacts the visibility and distribution of the research. Consistent adherence to these directives ensures that each work aligns with ACM's operational parameters and facilitates proper metadata extraction for digital library inclusion.

Implications and Speculations for Future AI Developments

While the document primarily functions as a technical guide, the advancements in template standardization suggested therein hold significant implications for AI-driven publication processes. Consistent template usage enables more efficient algorithmic parsing and indexing, thus enhancing AI’s ability to manage vast repositories of research effectively. Looking forward, the continuing refinement of submission standards could pave the way for fully automated systems for manuscript review and correction, leveraging AI to ensure adherence to formatting norms and semantic data extraction.

Conclusion

The ACM consolidated article template represents an essential evolution in academic publication, harmonizing submission formats across journals and conferences. This standardization facilitates both authors' submission processes and the seamless integration of their work into ACM's digital library ecosystem. The foresight in design ensures robust metadata implementation and rights management, critical aspects that influence accessibility and copyright adherence in digital publishing. Future developments in AI can further capitalize on these structured frameworks to enhance automated research dissemination and management.