Metalinguistic Information Extraction: Exploiting Metalanguage in Text

This work presents an empirical study of the use and function of metalanguage in expert scientific knowledge and special-domain languages, with special focus on how each field's terminology is established, modified and negotiated within the group of experts. Through discourse statements called Explicit Metalinguistic Operations (EMO), the dynamic nature of conceptual structures and the sublanguages that embody them are formalized and analyzed. On the other hand, it presents an implementation of a system for automatic extraction of metalinguistic information from specialized texts. The Metalinguistic Operation Processor (MOP) system extracts metalinguistic statements and definitions from special-domain documents, using finite-state machinery and machine-learning algorithms. The system creates semi-structured databases called Metalinguistic Information Databases (MID), useful for specialized lexicography, Natural Language Processing, and the empirical study of scientific knowledge.