Page Suggestions is a SciDEX feature that surfaces candidate wiki pages or entity topics for creation or expansion based on knowledge graph gaps, literature trends, and user engagement patterns.
Knowledge Gap Detection
Automated page suggestion systems derive candidate topics from multiple signal sources: unlinked entities in the knowledge graph that lack dedicated wiki pages, high-frequency search queries returning no results, literature-mined protein-gene-disease relationships not yet represented in Atlas, and strategic research priorities identified by Senate governance [@PMID:40298230]. The gap detection approach mirrors the landscape-gap framework used in Agora debates, where knowledge gaps are identified by analyzing which entities or relationships lack sufficient supporting evidence to form scored hypotheses.
The integration of literature trend analysis into page suggestion algorithms helps prioritize content creation for emerging research areas where the volume of recent publications suggests a rapidly evolving evidence base [@PMID:35504897]. In neurodegeneration, this might surface pages for newly validated genetic risk loci or for therapeutic approaches gaining traction after positive clinical trial results. This approach ensures that Atlas wiki content remains current with the scientific frontier rather than lagging behind published literature.
User Engagement Signals
...
Page Suggestions (SciDEX Feature)
Page Suggestions is a SciDEX feature that surfaces candidate wiki pages or entity topics for creation or expansion based on knowledge graph gaps, literature trends, and user engagement patterns.
Knowledge Gap Detection
Automated page suggestion systems derive candidate topics from multiple signal sources: unlinked entities in the knowledge graph that lack dedicated wiki pages, high-frequency search queries returning no results, literature-mined protein-gene-disease relationships not yet represented in Atlas, and strategic research priorities identified by Senate governance [@PMID:40298230]. The gap detection approach mirrors the landscape-gap framework used in Agora debates, where knowledge gaps are identified by analyzing which entities or relationships lack sufficient supporting evidence to form scored hypotheses.
The integration of literature trend analysis into page suggestion algorithms helps prioritize content creation for emerging research areas where the volume of recent publications suggests a rapidly evolving evidence base [@PMID:35504897]. In neurodegeneration, this might surface pages for newly validated genetic risk loci or for therapeutic approaches gaining traction after positive clinical trial results. This approach ensures that Atlas wiki content remains current with the scientific frontier rather than lagging behind published literature.
User Engagement Signals
Page suggestion systems also incorporate user engagement signals from the SciDEX platform. When researchers frequently visit related pages without finding a dedicated article, or when debate participants reference entities that lack wiki coverage, these signals feed the suggestion pipeline [@PMID:40776057]. The multi-dimensional quality scoring used in knowledge graph assessment applies equally to page suggestion prioritization, with novelty, evidence strength, and strategic relevance weighted against each other to generate ranked suggestion lists.
The suggestion pipeline supports both automated content skeleton generation and full content expansion by human curators or AI agents. The skeleton-first approach ensures that even low-priority entities receive minimal coverage, while the expansion pathway activates when quality reviewers determine a topic warrants full substantive treatment. This two-tiered approach balances coverage breadth against the limited human and AI curatorial capacity available for content creation.
The gap detection framework applies established principles from scientific discovery literature, which identifies knowledge gaps as absences of expected causal mechanisms or missing evidence chains linking established disease pathophysiology to therapeutic outcomes. By formalizing gap detection as a structured analytical task rather than an ad hoc curation exercise, page suggestion systems can scale to cover the full breadth of neurodegeneration research without requiring manual prioritization of every content creation opportunity.