DOMAIN SPECIFIC NATURAL LANGUAGE NORMALIZATION

摘要

A method for the domain specific normalization of a corpus of text including an industrial, organization, demographic or geographic domain. The method includes loading a corpus of text in a memory 310 of a computer and determining a domain for the corpus of text 320. The method also includes retrieving a lexicon of replacement words 330 for the determined domain. The method includes text simplifying the corpus of text using the retrieved catalogue of words 340. The domain may be determined through inference based upon words already present in the corpus of text. The domain may also be determined based upon meta-data provided. The list of replacement terms may be a set of source terms which can be mapped to one of a multiple different replacement terms which have a complexity value aligned with an average complexity score for the multiple different replacement terms.