Proposition Bank

Parse the question using a syntactic parser.

Identify the predicates and their arguments using PropBank annotations.

Map the semantic roles to potential answer candidates.

Rank the answer candidates based on semantic relevance.

Process the text using a named entity recognizer.

Identify predicates and arguments with PropBank.

Extract relationships between entities based on semantic roles.

Store extracted information in a knowledge base.

Parse the source sentence and identify predicates and arguments.

Map semantic roles to equivalent roles in the target language.

Generate the target sentence while preserving semantic relationships.

Evaluate translation quality using semantic similarity metrics.

Parse the document and identify predicates and arguments.

Score sentences based on the number of important semantic roles they contain.

Select the top-ranked sentences to form the summary.

Refine the summary for coherence and readability.

Parse the user's utterance and identify predicates and arguments.

Maintain a semantic representation of the dialogue history.

Generate a response that aligns with the semantic context.

Evaluate the response for semantic coherence and relevance.