Context-Aware Query Interpretation

Now & Then searches are based on context and concept, rather than key words. This means that the most relevant information found in response to your search need not contain a single word used in your Query.

Now & Then uses mathematical models of your word attributes and knowledge maps (ontologies), to interpret your queries and determine how they relate (conceptually and contextually) to your journal information. Query interpretation is iterative and not necessarily sequential.

Now & Then Query interpretation processes involve:

• iteratively "parsing" each query (both sentence and template queries) to determine search intent, search concept, and search context, resolve ambiguities, and rank relevance,

• mapping A "core" Query into the term attribute space,

• matching of Query content and context to your "facts and knowledge" ontologies and "maps" of your worldview,

• determination of time, date, and calendar attributes and connections between your Query and journal information (Intelligent  Calendar),

• Bayesian probability ratings for the most likely "correct" search outcomes for individual, combined, and iterative Query interpretation methods,

• requests for user guidance (using system suggested alternatives) if ambiguity cannot be automatically removed, and

• examination of retrieved information to identify relationships, structure, patterns, and sequence―to determine the most appropriate presentation and display methods.

Also see Product Features:
Search & Query   ▪   Language indexing & Search   ▪   Knowledge-based indexing & search

home