2006 Fiscal Year Final Research Report Summary
Study of the Visualization of Web Access Pattern for the Assisting of School Network Operation
| Project/Area Number |
16500591
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Educational technology
|
|---|
| Research Institution | The University of Tokushima |
|---|
Principal Investigator |
KANENISHI Kazuhide The University of Tokushima, Center for Advanced Information Technology, Assistant professor, 高度情報化基盤センター, 助教授 (80204577)
|
|---|
| Project Period (FY) |
2004 – 2006
|
| Keywords | Mining / SOM / Vizualization / User Behavior Pattern / Vector Space Model / Property Vector / Assisting of Network Operation / Administrator Assistance |
|---|
| Research Abstract |
The purpose of this research is to development of the system that assists the school network operation. This system is able to visualize the information of the Web access pattern in an elementary school, junior high school and so on. So, Visualization assists the detection of the anomaly access. The large part of the internet use in the elementary school is browsing of the web site. There is a possibility that undesired internet access to the outside is hidden. The ability that administrator knows the situation accurately is required. As she/he takes careful to the individual privacy, the operation process is difficult. Then, we propose the method that web browsing history is statistically processed, and the browsing history is visualized at the same time. We propose the use of the vector space model. The property vector is made from the browsing pattern. We use SOM (Self-Organizing Maps) to classify this property vector. As a result, the two dimensional map is obtained. This visualized information made becomes a map that shows the overhead view about the browsing pattern as a whole. The administrator comes to be able to understand the usage pattern by intuition by the visualization. This system is a type of web mining system This system extracts features of user interests from the web content that users browses, and after that, creates a feature map by clustering and visualizing We made the prototype system. The prototype system was tried for a certain period in an educational institution. As a result, the effect of the information visualization that we proposed was confirmed.
|
Research Products
(26 results)
