| Project/Area Number |
12680394
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Intelligent informatics
|
|---|
| Research Institution | Hiroshima City University |
|---|
Principal Investigator |
KITAKAMI Hajime Hiroshima City University, Information Sciences, Professor, 情報科学部, 教授 (50234240)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
KUROKI Susumu Hiroshima City University, Information Sciences, Associative Professor, 情報科学部, 助教授 (20225288)
FUJI Ren Tokushima University, Engineering, Professor, 工学部, 教授 (20264947)
|
|---|
| Project Period (FY) |
2000 – 2002
|
| Project Status |
Completed (Fiscal Year 2002)
|
| Budget Amount *help |
¥3,100,000 (Direct Cost: ¥3,100,000)
Fiscal Year 2002: ¥800,000 (Direct Cost: ¥800,000)
Fiscal Year 2001: ¥900,000 (Direct Cost: ¥900,000)
Fiscal Year 2000: ¥1,400,000 (Direct Cost: ¥1,400,000)
|
|---|
| Keywords | Constraint Database / Ordered Tree / Reconciliation / Heterogeneous Tree / Genetic Algorithm / Taxonomy / Parallel Processing / Depth-First Search / 調停グラフ / 階層グラフ / 異種木構造 / クラスタ交換法 / ビジュアル化 / ビジュアルデータベース / 系統樹 |
|---|
| Research Abstract |
This is a study of heterogeneous databases, which includes structures such as the taxonomic tree and conceptual hierarchy. In a study of effective constraint-query processing in two heterogeneous tree databases, we obtained the following results :
(1) A constraint query is represented by the selection of a subtree from each database and a minimum of computational crossover for a hierarchical graph constructed using two subtrees selected from two databases. Moreover, the constraint query, which is represented as the minimum of computational crossover, is different from the existing spatial query, which is a kind of constraint query that is processed by geometrical reasoning in spatial databases.
(2) In order to minimize the computational crossover for the graph, an interconnection matrix defined by us is constructed with a constraint between two leaf layers in two subtrees, and the matrix approaches a unit matrix from its initial status using matrix transformations. We found that our depth-first search is useful to approach the unit matrix. The depth-first search does not allow creating any crossover among the branches of each subtree.
(3) We designed a user interface implemented in JAVA, to visualize the result of the constraint query. We found that the interface is easy to use for the results of constraint query.
(4) In order to achieve faster computation to find minimum crossovers, we considered using a genetic algorithm. As a result, combining the depth-first search with the genetic algorithm was useful for the improvement of computational speed.
(5) A very large database is required to enhance the processing of a constraint query. We considered parallel and distributed processing for a. depth-first search on multiple computers. As a result, we obtained good performance on the computers.
|
