A Text Classifier About High Blood Pressure Based On Naive Bayes

With the quickening pace of life, people living pressure increasing, moreand more chronic diseases which frequently appeared in the elderly also appearin the adolescent population, as a major risk factor for cardiovascular disease,hypertension disease escalated into the key issues in medical research. With theadvent of the era of information, the rapid development of Internet technology,network reflect a large amount of information resources, and presents thegeometric form of growth, the world wide web development into themainstream of the world’s information exchange and sharing platform, peoplepublished information resources, gain information resources, learning from eachother, mutual communication on the Internet. People are more willing tounderstand the related medical information as soon as possible, via the Internetto quickly and effectively obtain needed to focus before treatment.A large number of text messages are stored in the swelling development ofInternet information, hypertension in high blood pressure after class textinformation collection, the texts are hypertension class information, these textsstill have a large amount of data and information query inconvenientshortcoming. Application of automatic text classification technology can increase the speed of information extraction, rapidly implement of text category,at present about text categorizations are mostly general text classifier,professional text classifier has not been very wide development in the field,also does not have the information mainly for hypertension text classifier. In thispaper, in order to solve the difficulties for selection of patients with high bloodpressure in information filtering, a kind of high blood pressure text classifier isput forward.Firstly the key technology of text categorization system is expounded,including the representation of Chinese word segmentation, text informationand text classification, text feature selection algorithms, focusing on theclassification principle of naive bayes algorithm. Mainly establishedhypertension information dictionary, its application in the classifier in Chineseword segmentation and the process of text dimension reduction, through thecombination of information gain and the high blood pressure informationdictionary feature selection methods，the establishment of the dictionary givesfull consideration to the importance of hypertension vocabulary; the corpus ofthe text categorization about hypertension is established, based on the Internetto collect a large number of texts, establish hypertension classification corpus;naive bayes classification algorithm principle is analyzed in detail, itsapplication in the process of hypertension text classification, through theexperimental study of the effect of the naive bayes classification; Aimed at thelimitation of the naive bayes classification, this paper puts forward the improved weighted naive bayes and its application in text categorization of hypertension,verified by the experiment, the classification effect is significantly increased, toachieve the goal of practical application;The work target is to study how to utilize the naive bayes algorithm for textclassification of hypertension and improve the efficiency of text classification ofhypertension in this paper. At the same time, this paper has researchfulcharacteristics. Thus, the system proposed in the paper may have somedeficiencies.