System development of Decision Support System for Media Planning
Project/Area Number |
16510129
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Single-year Grants |
Section | 一般 |
Research Field |
Social systems engineering/Safety system
|
Research Institution | Kwansei Gakuin University |
Principal Investigator |
IGAKI Nobuko Kwansei Gakuin University, Policy Studies, Professor, 総合政策学部, 教授 (40151253)
|
Co-Investigator(Kenkyū-buntansha) |
NAKAGAWA Yuji Kansai University, Informatics, Professor, 総合情報学部, 教授 (60141925)
ISADA Yuriko Tezukayama University, Business Information, Associate Professor, 経営情報学部, 助教授 (00351867)
|
Project Period (FY) |
2004 – 2005
|
Project Status |
Completed (Fiscal Year 2005)
|
Budget Amount *help |
¥1,700,000 (Direct Cost: ¥1,700,000)
Fiscal Year 2005: ¥600,000 (Direct Cost: ¥600,000)
Fiscal Year 2004: ¥1,100,000 (Direct Cost: ¥1,100,000)
|
Keywords | Media Planning / Optimization Algorism / Decision Support System / Knapsack Problem / Newspaper Advertisement / メディアプランニング / 最適化アルゴリズム |
Research Abstract |
In this research we attempt to develop a decision support system to solve exposure optimization problem of media planning. We focus on newspaper advertisement with a national-wide scale. Newspaper is the largest mass advertising media next to television. First we target four national newspapers, then add 44 local newspapers in Japan. If you make this newspaper problem in natural way, you will see a knapsack problem with complicated constraints which cannot be solved. We overcome this difficulty to divide this main problem into 16 partial problems which are knapsack problems with one constraint respectively. In general it is difficult to get an exact solution for knapsack problem with one constraint. However, we could get the exact solutions using HOPE (Hybrid Optimization Process Equipment) which is developed by Yuji Nakagawa who is one of members on our research team. Usually, for this kind of media planning problems, only approximate solutions are obtained. Therefore, the exact solutions we got are supposed to be great results. For example, we can verify the previous approximate solutions measuring the variation from the exact solutions respectively.
|
Report
(3 results)
Research Products
(6 results)