Dissertation Information for Jun Min Jeong NAME: - Jun Min Jeong
DEGREE:
- Ph.D.
DISCIPLINE:
- Library and Information Science
SCHOOL:
- Case Western Reserve University (USA) (1985)
ADVISORS: - William Goffman
COMMITTEE MEMBERS: - Blaine Garfolo - Thomas Gleason Morris - Tefko Saracevic
MPACT Status: Fully Complete
Title: The ecology of the scientific literature and information retrieval
Abstract: The emphasis of this research deals with the problems encountered in designing systems for more efficient and effective information retrieval used in the proliferation of literature. This research was designed to develop and test 1) the partitioning a large bibliographic data base into quality oriented subsets (qualify filtering), and 2) a system for effective and efficient information retrieval within subsets of data base (relevance). In order to accomplish this partitioning, the 'kernel' technique of graph theory was applied. In addition, a method of quality filtering utilizing the 'epidemic' theory and the 'obsolescence' of scientific literature was developed.
Quality filtering
It is assumed that scientific papers are eventually less valued and less used with the passage of time obsolescence), and that the transmission of ideas within a population will be treated as if it were the transmission of an infectious disease (epidemiology). The underlying assumption of quality filtering is that highly qualified papers will be identified according to their epidemic curve normalized by the characteristics that the scientific paper would be obsolete with time.
In applying a quality filtering method to the test collection, it was shown that the value generated from the differential equation of epidemic process could be an indicator for quality evaluation. As mentioned in the hypothesis, the highly qualified papers were identified according to their epidemic curve and normalized by the characteristic of obsolescence.
Relevance
The second hypothesis of this research was: Highly associated documents can be grouped and these groups can be represented by kernel sets which characterising the property of each clusters. Also, at the lower threshold values, the kernels can be reculstered together and hierarchically create new upper level kernel sets.
The characteristics of a kernel are defined in terms of independence and domination. In a connected graph, the elements in a kernel set should be independent of each other and these elements should dominate the other elements in the graph. That is, any two elements in a kernel may have no relationship directly, but they can be related to each other through the other, elements which are outside the kernel set. Hence, the elements in kernel set can represent the total graph.
In this study, the results showed that the kernel technique was very useful in the hierarchical structure of a cluster-based retrieval system because of the characteristic of representation.
| |
MPACT Scores for Jun Min Jeong A = 0
C = 0
A+C = 0
T = 0
G = 0
W = 0
TD = 0
TA = 0
calculated 2008-01-31 06:01:25
Advisors and Advisees Graph
|