臺灣博碩士論文加值系統

在科技日新月異的今日,新產品是企業的命脈所在。對網路產業業者
而言,推出新產品的成敗關乎企業之生死存亡甚鉅。過去的研究顯示新產
品的進入時機會影響產品的績效,有許多研究顯示先導者具有市場佔有率
的優勢,但不能一概而論；以權變理論的觀點而言,管理者必須將市場策略
及市場變數列入考慮。本研究包含： 1. 新產品進入策略模式：特別
強調產品定位的重要性。 2. 進入時機的決策變數：本研究歸納成產
品生命週期、市場競爭情形、公司競爭優勢、產品特性等四大類的決策變
數。但是從個案公司的問卷調查資料分析中顯示：除市場競爭情形中的「
潛在市場」因素外,其它的因素和進入時機都無顯著關係。分析其原因可
能為國內網路產業廠商以OEM為主,而且競爭優勢在「降低成本」,新產品
「何時」進入市場並非重點,更重要的是「如何」進入。 3. 用模糊邏
輯呈現知識及推理規則的知識庫系統雛型。系統中包括：根據訓練範例產
生推理規則的自動推理系統及總結多人對產品生命週期不同看法的方法。
由於無法取得適當且足夠的資料,因此無法驗證系統的有效性。 4. 根
據黑板架構原理設計成可分散處理專家知識的黑板資料區雛型。 最後
必須說明的是：就「新產品進入時機」這個主題而言,本研究屬於探索式
研究,需要更深入的研究才能得出選擇進入時機的一般模式。屆時,本知識
庫系統才有其實用的價值。

New products are the life-force of a company during an age of
quicktechnology change. For the network industry , there is no
exception. According to past researches , the timing for a new
product coming into the market is one of the major reasons for
its success or failure. Several studies show that earlier entry
is associated with greater market share, but in the contingency
perspective, managers should take the market strategy and
marketplace variables into account. In this paper we propose:
1. A model of new product's entry strategy. 2. A set of
decision variables of entry time. These variables are classified
into four groups: (1)new product's life cycle; (2)the
competition condition; (3)the company's competitive advantages;
(4)the product's features. In the analysis from a case company ,
we find that besides the only factor "potential market " which
belongs to the group of competition condition, the other factors
have little connection with entry time. This may be explained by
the fact that most new products of the case company are
primarily OEM and the company's competitive advantage is cost
leadership, so the entry time of new products is not the main
decisive factor ; rather, the way how new products are getting
into market sounds more important. 3. A knowledge-based
prototype by using fuzzy logic for the presentation and
inference. Besides, we also implement an automatic reasoning
algorithm which can infer many reasoning rules from past
experiences. Regarding the problem of the prediction of new
product life cycle curve, we use SAM(Similarity Aggregation
Method) to gather different experts' opinions. Because we don't
have enough data we can't verify the validity of this system.
4. A DAI(distributed AI) approach to resolve different views of
a group of experts. We design the structure of this system on
the basis of the concept of blackboard architecture. In
conclusion, we must point out that it requires more and deeper
empirical research for entry strategy to realize this knowledge-
based prototype.