1. Untitled
1.1. search completion
1.1.1. implement shingle-two autocomplete search
1.1.1.1. terms aggregation
1.1.1.1.1. def analyzer
1.1.1.1.2. def mappings
1.1.1.1.3. def queries
1.1.1.2. completion suggester
1.1.1.2.1. benefits
1.1.1.2.2. drawbacks
1.1.1.2.3. internal structure
1.1.1.2.4. processes
1.1.2. drawbacks
1.1.2.1. heavy aggregation over large text fields tax search engine
1.1.2.2. by default this method returns completion suggestions ordered from most commonly occurring to least commonly occurring
1.1.3. challenges
1.1.3.1. users expect search completions to return shortly after beginning typing
1.1.3.2. highly relevant
1.2. corrections
1.2.1. post search suggestion
1.2.1.1. mapping requires a field specifically set for suggestion
1.2.1.2. def a mapping
1.2.1.2.1. Untitled
1.2.1.3. pull back suggestions
1.2.1.3.1. Untitled
1.2.1.4. retrieve post-search suggestions in the context of a user's search
1.2.1.4.1. collation
1.2.1.4.2. Untitled
1.2.1.5. returned suggestions
1.2.1.5.1. Untitled
1.3. highlight
1.3.1. basic highlighter
1.3.1.1. must reanalyze documents in order to find the location of the matching terms within the document
1.3.1.1.1. minuscule for small docs
1.3.1.1.2. performance hit for large docs
1.3.2. postings / fast vector highlighter
1.3.2.1. benefits
1.3.2.1.1. avoid query-time processing
1.3.2.2. drawbacks
1.3.2.2.1. increased index size
1.3.2.3. posting highlighter for natural language text
1.3.2.4. fast vector highlighter can highlight matching terms independently
1.3.3. sample
1.3.3.1. queries
1.3.3.1.1. Untitled
1.3.3.2. partial results
1.3.3.2.1. Untitled
1.4. faceted search
1.4.1. query
1.4.1.1. Untitled
1.4.2. result
1.4.2.1. Untitled