Short answers, pulled from the story.
A semantic network is a knowledge base that represents semantic relations between concepts as a graph, with vertices standing for concepts and edges carrying the meaning of the relationship between them. It can be directed or undirected and is often expressed using standardized semantic triples. Semantic networks are used in natural language processing, knowledge representation, and information retrieval.
Richard H. Richens of the Cambridge Language Research Unit implemented the first computer-based Semantic Nets in 1956 as an interlingua for machine translation. Robert F. Simmons and Sheldon Klein independently developed their own implementation around 1962-1964, inspired by Victor Yngve. M. Ross Quillian and others at System Development Corporation also contributed through the SYNTHEX project in the early 1960s, and most modern use of the term traces back to publications from that organization.
Knowledge graphs are semantic networks with an added constraint: edges must come from a limited set of predefined relation types, which allows algebraic operations on the graph. The term was coined for a joint project by the Dutch universities of Groningen and Twente in the late 1980s. Over subsequent decades the distinction between the two became blurred, and in 2012 Google applied the name Knowledge Graph to its own search product.
WordNet is a lexical database of English that groups words into synonym sets called synsets and records semantic relations such as meronymy, holonymy, hyponymy, hypernymy, synonymy, and antonymy. It is one of the most studied examples of a semantic network. Researchers examining WordNet alongside networks derived from Roget's Thesaurus and word association tasks found that all three exhibit a small-world network structure.
The earliest documented use is the Greek philosopher Porphyry's third-century AD commentary on Aristotle's categories, which employed directed acyclic graphs as a mnemonic tool for representing how categories of concepts nest inside one another.
Semantic networks are used in natural language processing tasks such as semantic parsing and word-sense disambiguation. They are also used to analyze large text corpora to identify themes and reveal biases, to map research fields, and in information retrieval applications including plagiarism detection. Embedding techniques that place semantic network data into low-dimensional spaces support social network analysis and relationship extraction.