Documentation

Smart Select Filters

Smart Select uses AI-powered algorithms to automatically choose the most relevant papers from your search results. Instead of manually reviewing hundreds of papers, Smart Select narrows your evidence base to the studies that matter most for your specific research question.

All Plans
Top Semantic Relevance

Automatically select the papers most semantically relevant to your query using AI embeddings. This mode ranks papers by how closely their content matches the meaning of your research question, not just keyword overlap.

  • Uses vector embeddings to measure conceptual similarity
  • Captures papers using different terminology for the same concepts
  • Available on all plans including Basic
All Plans
Biologically Correlated

Select papers that share biological connections through the research knowledge graph. This mode surfaces studies linked by shared genes, Gene Ontology terms, MeSH descriptors, and domain-specific keywords — finding biologically meaningful connections that text similarity alone would miss.

  • Leverages the biology-native knowledge graph
  • Finds papers connected through shared pathways and gene networks
  • Reveals mechanistic relationships across studies
All Plans
Active Research Frontiers

Identify and select papers from emerging, accelerating areas of research. This mode focuses on clusters where a high proportion of publications are recent, surfacing the cutting edge of your field.

  • Prioritizes papers from rising research clusters
  • Uses temporal momentum analysis to detect accelerating areas
  • Ideal for staying current with the latest developments
How Smart Select Works

After running a search, Smart Select helps you build your evidence base efficiently:

  1. Run your search: BioSkepsis retrieves relevant papers from 40M+ publications
  2. Choose a Smart Select mode: Select the filter that matches your research goal
  3. Review selections: Smart Select highlights the recommended papers
  4. Refine manually: Add or remove papers before proceeding to analysis

When to Use Each Mode
  • Top Semantic Relevance: Best for focused queries where you want the closest conceptual matches to your question
  • Biologically Correlated: Best for mechanistic exploration where biological relationships matter more than keyword similarity
  • Active Research Frontiers: Best for staying current, identifying hot topics, or writing grant proposals that need to reference cutting-edge work