Technical terminology for domain specification and content characterisation

Abstract

The identification and extraction of technical terms is one of the better understood and most robust natural language processing (NLP) technologies within the current state of the art of language engineering. What is particularly interesting here is the clear understanding how to derive, from their linguistic properties, computational procedures for reliable identification and extraction of terms from technical, scientific, prose. In generic information management contexts, terms have been associated both with procedures seeking to identify a term set which uniquely distinguishes a document within a nearly homogenous document collection, and with procedures seeking to extract a representative terms sample which uniquely characterises a document's content. There is a wide range of uses for terminology, commonly identified with e.g. text indexing, computational lexicology, and machine-assisted translation; most of these employ the notion of terminology being representative of a given domain. This paper discusses some specific extensions of the terminology identification technology to make it fully capable of domain specification; it also presents extensions of the technology beyond domain specification, to the purpose of document characterisation. These extensions make terminology identification the foundation of an operational environment for document processing and content characterisation and abstraction; more generally, it becomes an immensely empowering technology in the age of growing information overload.