Text Classification In The Information Service Of The Government Aged-care Application Research

With the further development of the Internet and computer technology,as well as the social aging population increase,especially the government has published issused a wide range of aged-care policies and related information through its website and related Internet resources,people are paying more attention to how to use modern information technology to improve the social service capacity for aged-care.The aged-care information from government is authoritative and political,which avoids the information asymmetry caused by nongovernmental information's regional deviation,poor reliability and redundancy.Meanwhile,government news is made up of a large amount of unstructured Chinese text information.It has become an important subject with not only the scientific research value but also the potential application to provide valuable information services for the elderly,aged-care served and other relevant people through analyzing the old-supporting information in governmental news and policies.In order to extract valuable information from the government news,this dissertation uses the crawler technology based on Python to design the crawler module to the crawl the news and official documents on the government websites and store it in the local database,establishing a set of text training samples by manual collection and mark.The optimal parameters of the algorithm are selected by experiments and the performances on the text training set of three classification algorithms commonly used in binary classification are compared.After preprocessing the text,we select an effective feature selection method to reduce the feature dimension,and then use the classification model with the best comprehensive performance to filter the data collected in the database,getting the aged-care information.in the government news.Finally,Numpy and Matplotlib technologies are used to process the classification results,which will generate a series of visual experimental analyzable pictures.The experimental results show that it is feasible to use text categorization technology extracting valuable information from governmental aged-care information in the process from data collection to outcome analysis and processing.This approach will make up for gaps and deficiency in the field of aged-care services,and it has an positive effect on improving the social old-supporting service level when exploring furtherly the area of aged-care from the informationalized perspective.