Documentation
- Query Refinement
- Research Lenses
- AI-Powered Semantic Search
- Unbiased Computational Analysis
- Research Landscape
- Research Landscape Synthesis
- Trends and Momentum
- AI Research Assistant
- Citation-Grounded Answers
- Hypothesis Generation
- Methodology Generation
- Network Analysis
- Foundational Papers
- Hub Papers
- Bridge Papers
- Novel Leads
- PDF Export
Documentation
Trends and Momentum
Understand the temporal dynamics of your research field. BioSkepsis analyzes publication patterns to identify rising, stable, and declining research areas, helping you discover emerging frontiers and understand field evolution over time.
How Trend Analysis Works
Trend analysis examines publication years across your paper set to compute momentum metrics:
- Year Distribution: Analyzes when papers in each cluster were published
- Recent vs. Older Ratio: Compares papers from the last 3 years against older publications
- Median Year Calculation: Determines the central tendency of publication dates
- Momentum Classification: Assigns trend labels based on temporal patterns
Trend Labels
Papers and clusters are classified into three main trend categories:
60%+ recent papers: Most publications in this area are from the last 3 years, indicating active and growing research interest. These represent emerging frontiers and hot topics in the field.
20-60% recent papers: Consistent research activity over time with a balanced distribution between recent and older publications. These areas represent established research domains with ongoing interest.
Less than 20% recent papers: Fewer recent publications compared to older work, suggesting reduced research activity. May indicate mature areas, solved problems, or paradigm shifts to new approaches.
Momentum Classes
For more granular analysis, clusters are also assigned momentum classes based on the ratio of recent to older papers:
- Emerging: Only recent papers exist (ratio = infinity) - brand new research areas
- Accelerating: Ratio greater than 2.0 - rapidly growing interest
- Growing: Ratio between 1.0 and 2.0 - steady increase in activity
- Stable: Ratio between 0.5 and 1.0 - consistent research output
- Declining: Ratio between 0.25 and 0.5 - reduced activity
- Waning: Ratio less than 0.25 - significantly decreased interest
Filtering by Trend
Use trend filters in the Results Panel to focus on papers by their temporal characteristics:
- All: Show all papers regardless of trend
- Rising: Show only papers in rising/trending clusters
- Stable: Show papers in clusters with consistent activity
- Declining: Show papers in clusters with reduced activity
- Trending: Filter to emerging, accelerating, or growing momentum classes
Sorting by Momentum
Sort your results by trend momentum to prioritize papers based on research activity:
- Papers from rising clusters appear first
- Trend score combines recent paper ratio (70%) and median year normalization (30%)
- Higher scores indicate more active, recent research activity
Trend Over Time in Research Landscape
When generating a Research Landscape Synthesis, the AI includes a dedicated "Trend Over Time" section that provides:
- Global Temporal Trend: Overall field direction (increasing, stable, decreasing), year range, and median publication year
- Temporal Concentration: What fraction of papers are from the last 3 years, flagging potential recency bias
- Per-Cluster Momentum: Table showing momentum class, recent vs older paper counts, and momentum ratio for each cluster
- Field Evolution Insights: Interpretation of which areas are gaining or losing traction and what paradigm shifts may be occurring
Use Cases
Trend analysis helps you:
- Identify hot topics: Find emerging areas with accelerating research activity
- Understand field evolution: See how research focus has shifted over time
- Prioritize reading: Focus on rising areas for cutting-edge insights
- Identify gaps: Declining areas may represent opportunities for revival or new approaches
- Grant writing: Justify research directions by showing field momentum
- Avoid bias: High temporal concentration warnings help avoid over-reliance on recent work