A Study On Identification Rules Of Hedges From Biomedical Literature

Vagueness is a fundamental property of human language, since ancient times, Chinese andforeign scholars have never stopped exploring. In1965, Zadeh proposed the concept of fuzzy sets,marking the birth of fuzzy science. Then, a new discipline, fuzzy linguistics formed under thecombination of fuzzy theory and linguistics, which makes the study of vague language entered anew stage.Hedges, as a branch of fuzzy linguistics, once put forward it have caused widespread concern.Scholars discussed in many aspects and achieved fruitful results. The only drawback is thatprevious studies based on the actual data from the perspective of computer information processingand specific studies on identification rules of hedges is relatively small, so does it in the field ofbiomedicine. Based on the above, the paper constructs a hedges corpus of Chinese biomedicalfields, hoping to explore effective identification rules of hedges.Research on the classification of hedges have also been very in-depth, yet it will show somedeficiencies when these traditional classification method is applied to a certain type of genre orcomputer information processing. Therefore, we has made a new classification of Hedges inbiomedical literature, in which it is divided into two major categories of approximates and shields.In order to study identification rules of hedges, this paper produced a quasi Chinese hedgesword list according to the translation of English research. Then analyzed the words one by oneaccording to the classification above, and come to the final Chinese hedges word list frombiomedical scientific papers. During the analysis, we found that it is closely related to its parts ofspeech, meaning, syntactic component and other factors whether the language form is hedge or not.After discussed the rules of hedges, rules of hedges scope were also analyzed. It is mainly dueto the different syntactic constituents which caused by its parts of speech. Hedges, therefore,divided into adjectives, adverbs, conjunctions, verbs, approximates, parenthesis and classifierphrases, has been analyzed one by one. The classification and identification rules proposed in thispaper will promote the research on hedges, which is proved to be effective by language materials inthe hedges corpus constructed.