2005 Fiscal Year Final Research Report Summary
Development and Implementation of a System supporting the Epidemiological Analysis of Infectious Disease Information collected from the Internet Resources.
| Project/Area Number |
16500059
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Media informatics/Database
|
|---|
| Research Institution | Kobe University |
|---|
Principal Investigator |
MORISHITA Junya Kobe University, Faculty of Cross-Cultural Studies, Professor, 国際文化学部, 教授 (20182230)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
SHIMADA Masaaki Nagasaki University, Institute of Tropical Medicine, Professor, 熱帯医学研究所, 教授 (70124831)
NISHIYAMA Toshimasa Kansai Medical University, Faculty of Medicine, Professor, 医学部, 教授 (10192254)
KIYOMITSU Hidenari Kobe University, Faculty of Cross-Cultural Studies, Associate Professor, 国際文化学部, 助教授 (20304082)
|
|---|
| Project Period (FY) |
2004 – 2005
|
| Keywords | Internet / Infectious Disease / Analytic Epidemiology / Text Mining / Dynamic Classification / Database / Data Culturing / Scientific Database |
|---|
| Research Abstract |
In the Internet, huge and latest information can be acquired in every moment. If we could analyze the epidemiology of infectious disease information collected from the Internet resources, we might acquire a new epidemiological aspect different from a traditional one. To accomplish this aspect, we have developed a system that supports the epidemiological analysis of infectious disease information gathered from the Internet resources.
So far we had focused to the data of the ProMED mailing list as an epidemiological information resource, because in each mail the editor adds a special entry, from which we can easily retrieve the disease case classification, the field classification and the location information of the epidemiological event announced by the mail. At the first year, to remove the structural dependence to the entry, we examined the possible implementation of dynamic and automatic classification by the full text retrieval of the mail. As a result, we were able to achieve the sa
… More
me classification with the system without the editor's special entry.
At the second year, to expand the coverage of the system to more general Internet resources and to realize the system to be a more general dynamic classification system, the evaluation related to the maintenance of the classification tree data was done and the problems against the expansion of the coverage were also discussed.
The classification tree data is a composite object of classification items and retrieval keywords and it should be a growing database. We decided it to be a semi-structured data and discussed its implementation. The content of the research of this semi-structured data was able to make the result public with the content of some other fields. But it takes time to put enough amounts of the epidemiology and infectious disease terms in the system and to maintain the classification tree data. This has been left as a problem in the future.
To open it to the public, the prototype is being prepared. Especially, it is scheduled to make it as the Web application in the part of the visualization of the result of the retrieval so that it is possible to use it globally. Less
|
Research Products
(19 results)
