Home > Page Design > Algorithms > Topic Overlap Scores

Topic Overlap Scores: PageRank for Ideas

The Problem Google Solved (That We're Solving Again)

In 1998, Google faced a problem: when someone searches for "global warming," which of the millions of relevant webpages should appear first? Every page mentioning climate could claim relevance. But clearly a NASA climate science page deserves higher ranking than a blog post that mentions temperature once in passing.

PageRank solved this by measuring not just relevance, but strength of relevance. How central is this page to what the user actually wants?

The Idea Stock Exchange faces the exact same challenge. Consider the topic "Global Warming":

• 98% overlap: "CO2 emissions cause atmospheric warming" (this IS the core topic)
• 75% overlap: "Carbon taxes reduce emissions" (solution, not phenomenon)
• 25% overlap: "Individual recycling helps environment" (mentions climate, focuses on waste)
• 15% overlap: "Nuclear power is safe" (connects via energy policy, several steps removed)

Without systematic scoring, every belief remotely connected to climate floods the Global Warming page. Core scientific claims get buried under policy debates, individual action tips, economic analyses, and everything else that touches the topic tangentially.

Topic Overlap Scores solve this. They're PageRank for ideas—ranking beliefs by topical centrality instead of ranking webpages by link quality.

---

What Topic Overlap Actually Measures

The score (0-100%) answers one question: How much of this belief's meaning is specifically about this topic?

Example: "CO2 emissions from fossil fuels increase atmospheric warming"

Topic	Overlap	Why
Global Warming	98%	The belief is primarily about this
Fossil Fuels	60%	Mentions coal/oil but focuses on climate effect
Government Policy	12%	Implications exist but belief doesn't address policy
Individual Responsibility	5%	Tangential connection to personal consumption

On the "Global Warming" topic page, this belief appears at the top. On "Fossil Fuels," it appears prominently but lower. On "Government Policy," maybe page 3. On "Individual Responsibility," buried deep or filtered out entirely.

Different pages, different rankings, same belief. The overlap score determines where it appears.

---

How Scores Get Calculated

Here's where it gets elegant: the scoring methodology is itself scored using ReasonRank.

We don't decree that belief X has 75% overlap with topic Y. We build an argument structure where contributors propose scores and defend them with evidence.

The Meta-Structure: Arguments About Organization

For any belief-topic pairing, the system generates a question: "How strongly does [Belief] relate to [Topic]?"

This question gets its own argument tree:

Question: How much does "Government carbon pricing reduces emissions" overlap with "Global Warming"?

Arguments for HIGH overlap (75-90%):

• Primary purpose is addressing climate change mitigation (strong linkage to core topic)
• Policy interventions are major sub-category of climate discussions
• Keywords overlap significantly: "carbon," "emissions," "warming" (82% textual match)

Arguments for MEDIUM overlap (40-60%):

• Focuses on government action, not climate phenomenon (belongs primarily on "Climate Policy")
• Economic and political dimensions dominate content (multiple topic crossover)
• Assumes climate science as background without engaging directly

Arguments for LOW overlap (10-30%):

• Fundamentally about taxation and government power (spans many topics beyond climate)
• Someone reading only climate science might find policy prescriptions off-topic

Each argument gets evaluated using standard evidence scoring:

• Keyword/semantic analysis: 82% term overlap supports HIGH score
• User search behavior: 73% searching "global warming" also search "carbon tax" in same session supports HIGH
• Expert classification: Climate scientists categorize as "mitigation policy" (sub-category) supports MEDIUM-HIGH
• Citation patterns: Papers citing this belief also cite core climate science 68% of time supports MEDIUM-HIGH

Final score emerges from weighing competing arguments, exactly like a truth score emerges from weighing pro/con evidence.

In this case: ~75% overlap—high enough for prominent placement on Global Warming page, not so high that it crowds out core scientific claims.

---

Real-World Example: The Individual Actions Debate

One of the most contentious overlap questions in climate: Do beliefs about individual behavior belong on the "Global Warming" page?

Belief: "Individual recycling and conservation significantly reduce carbon emissions"

Arguments for HIGH overlap:

• Personal responsibility is major frame in climate discourse (45% of climate articles mention individual actions)
• Users searching "global warming" frequently ask "what can I do?"
• Directly addresses emissions reduction, core mitigation strategy

Arguments for LOW overlap:

• Individual actions account for ~25% of emissions; industrial/systemic sources dominate (EPA data)
• Belief focuses on behavioral psychology and personal ethics, not climate science
• Conflating individual actions with systemic solutions creates false equivalence (deflects from policy needs)
• Belongs primarily on "Environmental Ethics" or "Sustainable Living" topics

The linkage analysis reveals moderate connection. Individual actions do affect emissions, but the belief's emphasis on personal responsibility shifts focus away from atmospheric science that defines "Global Warming."

Result: ~40% overlap. Appears on Global Warming page but lower in ranking—perhaps page 2 or 3. Appears more prominently on "Climate Solutions" (65%) and "Sustainable Living" (85%).

---

Recursive All the Way Down

The beauty: every component has its own argument tree. Dispute a Topic Overlap Score? You're not arguing with a human editor—you're proposing a better scoring argument backed by evidence.

Level 1: The Belief
"CO2 emissions cause atmospheric warming"

Level 2: Topic Overlap Question
"How much does this overlap with 'Global Warming'?"

Level 3: Overlap Arguments

• Pro-HIGH: "The belief IS the core definition"
• Evidence: IPCC defines global warming precisely this way (primary source verification)
• Evidence: 98% of climate papers discussing warming address CO2's role (meta-analysis)
• Pro-MEDIUM: "It's about atmospheric chemistry, not warming specifically"
• Evidence: Also overlaps with "Atmospheric Science" at 75%
• Counter: Atmospheric science and global warming are inherently linked

Level 4: Evidence Verification
Each piece of evidence gets its own verification score, source quality assessment, linkage strength...

It's recursive reasoning—ReasonRank evaluating ReasonRank evaluating beliefs. The system bootstraps itself into increasingly accurate organization through the same mechanism it uses for everything else: proportioning conclusions to evidence.

---

Why This Prevents Gaming

Someone tries to inflate overlap scores for better placement? They'd have to:

1. Present verifiable evidence that belief's content substantially addresses the topic (keyword analysis, citation patterns, semantic relationships)
2. Defend against counter-arguments showing why belief belongs primarily elsewhere
3. Withstand algorithmic checks comparing proposed score against automated text analysis, user behavior data, expert classifications
4. Maintain consistency across all topic pairings (95% overlap with Topic A + 90% with Topic B requires mathematical coherence)

Gaming Topic Overlap Scores is like convincing Google your blog deserves #1 ranking for "climate science" when it mentions temperature once. The algorithm sees through it because all the evidence points the other direction.

---

Dynamic Reranking

Scores update automatically when underlying evidence changes.

Major study reveals individual actions account for 45% of emissions instead of 25%? The "individual recycling" belief's overlap with "Global Warming" increases proportionally. Belief rises in ranking—no human editor required.

New research shows CO2's warming effect amplified by methane feedback loops? Methane beliefs' overlap scores with "Global Warming" increase. Pure CO2 beliefs might decrease slightly as topic expands to include multiple greenhouse gases.

Organization evolves with understanding—exactly like scientific knowledge should.

---

User Controls: Customize Your View

Algorithm suggests 75% overlap. Great. But different users want different rankings:

• Strict mode: Only beliefs >80% overlap (core topic content only)
• Comprehensive mode: All beliefs >20% overlap (include tangential connections)
• Solutions-focused: Weight policy/action beliefs higher than scientific explanations
• Science-focused: Weight empirical claims higher than normative recommendations
• Controversy filter: Highlight beliefs with high disagreement regardless of overlap

Different users, different priorities. But underlying data structure (overlap scores + justifying arguments) remains consistent and auditable.

---

Integration with Other Scores

Topic Overlap works alongside other metrics for multi-dimensional ranking:

Final Ranking Formula for "Global Warming" Page:

Component	Weight	What It Measures
Topic Overlap Score	40%	How central is this belief to the topic?
Truth Score	35%	How well-supported is this belief overall?
Controversy Score	15%	How much active disagreement exists?
Recency Weight	10%	How recently has new evidence emerged?

Users can adjust weights: "Show me well-established core claims" versus "Show me cutting-edge debates."

A belief with 98% topic overlap but 15% truth score (thoroughly debunked core claim) might rank lower than a belief with 70% overlap and 92% truth (well-supported related concept). System balances relevance with quality.

---

What This Means in Practice

Before Topic Overlap Scores:

Every belief mentioning "carbon" floods Climate Change page. Core scientific claims buried under policy debates, individual action tips, economic analyses, conspiracy theories. Finding central thesis requires wading through hundreds of tangentially related arguments.

With Topic Overlap Scores:

"Global Warming" page opens with beliefs scoring 90%+ overlap—core scientific claims about greenhouse gases, temperature measurements, atmospheric physics. Scroll down: mitigation strategies (60-80%). Keep going: economic impacts, individual actions, political debates (30-60%). Everything findable but appropriately ranked.

It's the difference between a library where every book mentioning "Napoleon" is randomly shelved together, versus a library where biographies, military history, French history are clearly labeled, all accessible, but organized so you find what you're actually looking for.

---

The Meta-Lesson

Topic Overlap Scores reveal the fundamental principle underlying the entire Idea Stock Exchange:

Every decision about knowledge organization is itself a knowledge claim that requires evidence.

We don't trust human editors to decide what's relevant because humans have biases, blind spots, limited time. We don't trust pure algorithms because they lack nuance and context.

Instead, we create a transparent system where:

• Algorithms propose scores
• Humans provide evidence for adjustments
• Community verifies both
• Organization emerges from collective analysis, not individual decree

This is systematic knowledge architecture. Not centralized control. Not pure chaos. Structured emergence through evidence-based reasoning.

---

Why This Changes Everything

Right now, topic organization is either:

• Editor-controlled: Wikipedia decides what goes where (centralized, slow, subject to bias)
• Algorithm-controlled: Google/Facebook decide what you see (black box, optimizes for engagement not truth)
• Chaos: Reddit/Twitter threads where everything scatters chronologically

Topic Overlap Scores create a fourth option:

• Evidence-controlled: Transparent algorithms + community validation + automatic updates
• Every placement justified by verifiable criteria
• Every justification challengeable through new evidence
• Every score updating dynamically as knowledge improves

This is how we organize truth at scale. Not by trusting gatekeepers. Not by surrendering to algorithms. By making the organization process itself subject to the same systematic reasoning we apply to everything else.

---

Get Involved

We're building infrastructure for systematic knowledge organization. Tools that measure topical centrality. Systems that justify every placement decision. Frameworks that update automatically when evidence changes.

Ways to contribute:

• Evaluate overlap scores for major topics
• Build argument networks defending or challenging placements
• Test the ranking algorithms and report edge cases
• Create training materials for overlap evaluation
• Propose improvements to the scoring methodology

Join the project. Challenge our assumptions. Propose better frameworks. Help build the system.

Because here's what's at stake: Knowledge requires organization. Organization requires criteria. Criteria require justification. And justification requires evidence.

We can't fix human bias. But we can build systems where bias becomes visible, where criteria are transparent, where every organizational decision shows its work.

Topic Overlap Scores are how ideas find their pages. Evidence decides the hierarchy. The infrastructure sorts itself.

---

Related Scores Needed to be Calculated:

Argument scores from sub-argument scores

Evidence Scores

Book Logical Validity Score

Importance Score

Linkage Score Code

Truth Scores

Media Truth Score

topic overlap scores

Objective criteria scores

Media Genre and Style Scores