Document worth reading: “Towards Compositional Distributional Discourse Analysis”

Categorical compositional distributional semantics current a method to derive the meaning of a sentence from the meaning of its explicit individual phrases: the grammatical low cost of a sentence routinely induces a linear map for composing the phrase vectors obtained from distributional semantics. In this paper, we lengthen this passage from word-to-sentence to sentence-to-discourse composition. To acquire this we introduce a notion of major anaphoric discourses as a mid-level illustration between pure language discourse formalised in the case of major discourse illustration constructions (DRS); and data base queries over the Semantic Web as described by major graph patterns inside the Resource Description Framework (RDF). This affords a high-level specification for compositional algorithms for question answering and anaphora choice, and permits us to offer a picture of pure language understanding as a course of involving every statistical and logical sources. Towards Compositional Distributional Discourse Analysis