Building on the breakthrough natural language understanding capabilities of Wolfram|Alpha, the Wolfram Language has integrated features for accepting input in natural language and in other forms that require semantic understanding for interpretation. These features are important both in direct entry of Wolfram Language input and in calling on the Wolfram Language from APIs, forms, and other external and cloud constructs.
Interpreter — define how a string should be interpreted (directly or semantically)
SemanticInterpretation — interpret a string semantically with a particular target
Restricted — define a restriction (e.g. numbers within a certain range etc.)
DelimitedSequence — allow multiple objects in input, separated by delimiters
AmbiguityFunction — define how to handle ambiguities in semantic interpretation
AmbiguityList — representation of possible interpretations for ambiguous input
Failure — representation of failure to interpret as a particular type
Import — import from a file, in hundreds of possible formats
ImportString — import a string, using hundreds of possible file formats
SemanticImport — import structured data from a file semantically
SemanticImportString — extract structured data from a string semantically
ToExpression — interpret a string as Wolfram Language or similar input
GeoPosition — input geo positions in standard formats
FindGeoPosition — find geo positions from IP addresses, street addresses, etc.
BinaryRead — read binary data