A Study on Composition of Medical Examination Service by Semantic Network Analysis
Ⅰ. Introduction
Medical examination is a screening for detecting diseases hidden in the examinees and treating in the early stages. National medical examination of Kor ...
A Study on Composition of Medical Examination Service by Semantic Network Analysis
Ⅰ. Introduction
Medical examination is a screening for detecting diseases hidden in the examinees and treating in the early stages. National medical examination of Korea is composed of limited items, and a problem of selecting items of medical examination is raised due to the costs for private medical examination.
Accordingly, a new study methodology is necessary to prepare the basis of medical examination items for seeking the early detection of diseases and decline of prevalence rate and death rate while maximizing medical service consumers’ satisfaction.
Ⅱ. Purpose and Background of Study
Medical examination of Korea is divided largely into national examination provided by health insurance benefits and private examination that individuals pay costs as examination outside of the system. Although examination items are partially adjusted by age and gender in recent national examination, the basis of examination items is unclear in general. Therefore, systematic basis of examination items should be supported for implementation of examination customized to the characteristics of examinees. The basis of examination items for private examinations is also weak and evaluation is not properly implemented as well. Thus, this study is suggesting a study methodology for medical examination service that consumers can be satisfied and effect can be maximized clinically as a strategic medical service product. Through this study, First, a qualitative analysis on the medical examination desired by examinees who are the medical service consumers is conducted and Second, this study seeks a composition methodology for items of medical examination service that examinees are highly satisfied with.
Ⅲ. Contents and Methods of Study(Scope)
Contents of this study are as follows. First, this study considers a medical examination program by and large. Second, this study analyzes online health information. Third, this study calculates co-occurrence of refined data (morphemes) and produces a semantic network based on the foregoing.
Study methods used for proceeding with aforementioned study contents are as follows. First, after manufacturing a web crawler designed to collect necessary text information by itself, this study used text mining technology that classifies collected data as a unit of morpheme. Through the foregoing, this study extracted medical-related nouns from the texts. Second, this study conducted association analysis to draw kew words by finding a collocation from extracted morphemes and discover an association rule. Third, this study calculated co-occurrence of morphemes based on the discovered association rule, and framed a network on the basis of structure and strength of network connection between morphemes.
Ⅳ. Results of Study
Crawler conducted data collection targeting 1,002 cancer-related articles and 1,000 cases of counseling request medical service consumers left on the medical consulting site from January 13, 2003 through January 18, 2016. Collected articles and counseling requests were extracted as morphemes through natural language processing, and as a result of analyzing collocations based on the extracted morphemes, collocations such as ‘carcinogenic substance’, ‘uterine cervix’, ‘breast reconstruction’, ‘radiotherapy’ and ‘anticancer therapy’ were extracted from cancer-related articles, and collocations like ‘obesity diet’, ‘adipolysis’, ‘scoliosis’, ‘itchy anus’, ‘chronic fatigue’ and ‘examination result’ were extracted from medical consultation texts.
As a result of association analysis on the cancer-related articles, major cancers according to the gender had high percentage, and articles on the methods of therapy after cancer incidence were also extracted. Association of articles on the breast cancer turned out to be high as well. Association analysis on the medical consultation texts shows that many people have problems such as musculoskeletal disease, feedback on the result of medical examination, obesity, hemorrhoids etc.
As a result of semantic network analysis, medical examination related to cancer needs a particularized examination for different gender, and validity of introduction of intensive cancer screening items for medical service consumers in their 40s or over was shown. Meanwhile, results of semantic network analysis on the medical consultation showed the necessity of demand for medical examination items such as muscloskeletal disease and obesity in the adolescent period, and joint pain or respiratory disease due to aging etc.
Ⅴ. Expected Effect of Study Results and Application Plan
Academic contributions of this study are as follows. First, because it is a mechanical data analysis method, not hypothesis test using statistics, this study suggested a study methodology which is best fitted for studies for reflecting consumers’ demands. Second, interdisciplinary different views on the study results can be narrowed through a semantic network analysis. Third, an opportunity for convergence between studies that both medical and management are simultaneously considered is expected to be prepared. Fourth, medical service consumers’ request for consultation fall under the purview of the stage of attention or interest, so this study can be a good example that widens the scope of studies on consumer behavior.
In addition, this study can be utilized as follows. First, this study can be utilized as guidelines for the direction of studies on the medical management.
Second, the foundation for training human resources with practical competence desired by companies can be established using the methodology of this study. Third, this study can be utilized as a good medium for academic-industrial collaboration.