getting started in text mining part two开始在文本挖掘两个部分.pdfVIP

Message from ISCB Getting Started in Text Mining: Part Two 1 2 2 Andrey Rzhetsky , Michael Seringhaus , Mark B. Gerstein * 1 University of Chicago, Chicago, Illinois, United States of America, 2 Yale University, New Haven, Connecticut, United States of America tronic access to text obviously impacts all Because IE yields a collection of phrases stages of text mining. linking entities through predicates, one of Once the documents have been chosen its simplest but valuable uses is to answer by an information retrieval engine, a simple questions posed to an automated computer scans the text and picks out system [10]. In this approach, human the various entities (objects, concepts, and questions are digested by a linguistic symbols) in each sentence. This process, engine (likely using the same process as We are, in a sense, drowning in informa- called named-entity recognition [3], draws employed on original mined text) and tion. Today, it is unusual for scientists even upon dictionaries of synonyms and hom- mapped to simple phrases. These question to read a journal cover to cover—much onyms, in addition to machine-learning phrases are then queried against the less to personally