Ask Kalicube

Indexing Annotation Selection Filters form Level 3 of Jason Barnard's Indexing Annotation Hierarchy. These annotations route content chunks to the appropriate competition queue based on query characteristics: (1) Intent Match - does the chunk serve informational, transactional, navigational, or commercial intent? (2) Expertise Level - is the content beginner, intermediate, or expert level? (3) Claim Type - is the content a definition, process, comparison, recommendation, opinion, or fact? (4) Actionability - can the user act directly on this information? Unlike gatekeepers that eliminate, Selection Filters categorize. They don't remove chunks from consideration; they determine WHICH other chunks a given chunk competes against. Wrong pool placement means competing against better-matched content and losing.

Selection Filters are traffic directors. They don't kick you out - they tell you which line to stand in. Your expert-level content gets routed to the expert queue. Your how-to content goes to the procedural queue. Your definition goes to the informational queue. The danger? Wrong routing means wrong competition. Expert content served for a beginner query loses to actual beginner content. You didn't fail quality - you failed placement.

Search systems serve different content for different intents, even for similar queries. "Python" for a beginner gets tutorials; for an expert gets documentation. "Best CRM" for research gets comparisons; for purchase gets pricing. Selection Filters are how this routing happens at the annotation level. Understanding them explains why the same content wins for some queries and loses for others - it's not quality, it's category.