網路廣告媒體選擇之研究

A Study of Advertiser's Web Media Selection

媒體管理實務上，廣告主如何選擇最適的網路媒體以刊登廣告？此類決策問題，由於各種現象與決策程序的複雜性，造成單一決策者想要盡其所能考量所有問題構面，並做成有效的決策，愈來愈不容易，所以決策者，常邀請一群專家，參與方案的評選工作。因此這類管理問題，大都是由一個委員會（或評審小組）來選擇求解問題的方案，而非由一個人所決定，此類的決策問題屬於群體決策問題。其次，此類問題，具開放與非結構化的特性，決策常需納入複雜的環境因素，所以具多準則的特性，有些準則常是定性的，深受參與決策者的經驗與主觀判斷左右。另一特性是準則權重也常受決策者、準則與環境變異而改變。因此，這類管理決策問題本身充滿著模糊概念。本文旨在模糊環境下，建構一套網路廣告媒體選擇模式。本文針對1998年台灣地區十大網路廣告主、十大傳統媒體廣告主進行問卷調查，再依問卷利用因素分析，抽離出網路廣告主選擇網站之關鍵因素，進而建構出一刊登網路廣告網站之層級結構，文中提出一套模糊多評準決策方法的演算法則；該演算法則是根據三角形模糊數與語言變數兩個主要的觀念，作為評估準則之「重要性」與方案滿足各準則之「滿意度」評估之用。決策者可利用已設定好的語言值集，分別就準則的相對權重，與方案滿足準則之程度進行評分，並利用層級分析法求出關鍵因素之權重，經適當加權後，利用Chang and Chen的排序方法，求出適合刊登網路廣告之網站，以供廣告主決策之參考。文中以中國信託信用卡中心的網路廣告評選網站問題為例，對本文所提出之模式進行實證應用。

How to select web advertising media is a very complex process. It is a multi-criteria decision making problem in ill-structured situations. We must make a careful analysis among criteria, alternatives, weight, and decision makers before making a web advertising decision. Using the conventional crisp decision method, we always have to find precise data. But under many conditions, we cannot get precise data because the data are from the experience and the judgment of decision makers. The assessments of alternatives versus various criteria and the importance weight of criteria, however, often depend on judgment of approximation. Due to this type of existing fuzziness in decision making problem, we would expect a gradual transition from membership to nonmembership of an innovation's consonance with specified criteria as well as the relative importance weight of criteria. The purpose of this paper is to provide a multi-criteria decision making method to advertiser web media selection under fuzzy environment. This paper is aimed at the top ten web advertisers and the traditional media advertisers of Taiwan in 1998 to make an investigation. According to the results of the investigations we use the factor analysis to get the key factors of structure of web sites for posting web advertisements. This paper uses factor analysis to condense the advertiser web advertising media selection criteria first, to construct the hierarchical structure of web advertising media selection .An algorithm for evaluating web media selection is proposed in this paper. The algorithm is based on the concepts of triangular fuzzy number and linguistic values to assess the preference ratings of linguistic variable, 'importance' and 'satisfaction'. The subjects can employ an assumed rating set to show the individual perception of the linguistic variables. Then We use the Chang and Chen's ranking method to rank the final scores of web advertising site for web media selection. An example of the Chinatrust credit card center is discussed to see the possible advantages of the proposed method.