Semantic Parsing Task
- output: a Semantic Model (in a semantic representation language such as First-Order Logic).
- Performance Metric: Verifiability, Unambiguousness, Canonical Form, Inference, Expressiveness
- It can range from (typically) being a Natural Language Semantic Parsing Task to being a Computer Program Semantic Parsing Task.
- It can range from being a Shallow Semantic Parsing Task (such as extra-propositional semantic parsing) to being a Deep Semantic Parsing Task.
- It can be solved by a Semantic Parsing System (that applies a Semantic Parsing algorithm).
- See: Semantic Linguistic Analysis, Semantic Analysis Task, Sentence-level Semantic Analysis Task.
- (Liang, 2016) ⇒ Percy Liang. (2016). “Learning Executable Semantic Parsers for Natural Language Understanding.” In: Communications of the ACM Journal, 59(9). doi:10.1145/2866568
- Verifiability: With the representation scheme, it must be possible to compare (or match) the meaning of a sentence against the knowledge base.
- Unambiguousness: linguistic input may have several legitimate interpretations. A desired meaning representation should have the ability to tell which are more likely or unlikely
- Canonical form: It is desired that sentences with the same meaning should be assigned the same representation
- Inference: Inference refers to a system’s ability to draw valid conclusions based on the meaning representation ofinput and/or its store of knowledge
- Expressiveness: The expressiveness of a meaning representation language is a measure of the various meanings it candescribe.In principle, there is a very wide range of input and knowledge base.We want a meaning representation method that canaccurately represent any semantic natural language sentences.