Title page for etd-0729113-105349

URN	etd-0729113-105349	Statistics	This thesis had been viewed 586 times. Download 1 times.
Author	Zhi-Hao Zhong
Author's Email Address	No Public.
Department	Institute of Electrical Engineering
Year	2012	Semester	2
Degree	Master	Type of Document	Master's Thesis
Language	zh-TW.Big5 Chinese	Page Count	66
Title	An Analysis of Library Readers Using Clustering Coefficient
Keyword	Complex Network Associate rules Cluster Data Mining Data Mining Cluster Associate rules Complex Network
Abstract	In recent years, the quantity of data explodes. Under this circumstance, automated library system records reader’s information in the database. A lot of information is hidden in the data. Data mining is to discover the data and to turn it into information which can be used. 　Data mining has been used in many areas such as marketing and customer relationship management (CRM). Based on the result of data mining, the owners can change merchandise display and understand customers’ life style and habits in order to increase the volume of sales. To use data mining in libraries allows us to relocate the books and to recommend book lists to readers. In the long term effect, the book circulation can be improved. 　Many internet services could provide different types of personalization and customization. The technologies allow the service providers to give more specific service by understanding customer behaviors. However, this technology has been neglected in the libraries. Therefore the thesis will focus on readers’ reading history in the university and will try to discover the phenomenon among clusters. The research methodology will be data analysis and data mining, especially clustering analysis. 　In this thesis, clustering coefficient used to analyze behaviors of reader, and link weight are used to clustering reader into groups. It is convinced that non-exclusive hierarchical cluster research method can also be applied to the internet analysis. For example, it can be used to analyze the effect of different nodes in the internet and to figure out if the node is qualified to be the core node. Moreover, it is also useful to analyze the relationship in the social network.
Advisor Committee	Hui-Chu Chang - advisor Cheng-Chin Lin - co-chair Guo-Tong Lan - chair
Files	etd-0729113-105349.pdf indicate in-campus access at 1 years and off-campus access at 3 years
Date of Defense	2013-07-15	Date of Submission	2013-07-30

Back to Results