網站部落格之顧客口碑評論分析研究-以台灣咖啡飲料市場為例

網路已成為聚集消費者評論的極佳環境，越來越多的消費者使用網路來表達其個人對產品的看法，然而消費者發表的產品評論對企業而言是至關寶貴重要的資訊，因為這些訊息是有關於客戶對產品的關注與使用經驗。為了協助企業能夠確切掌握消費需求趨勢與脈動，本研究提出以擷取網路產品評價為基礎的研究架構，透過對網路資訊的偵測與蒐集機制，將所擷取的網路文章內容進行中文斷詞處理、特徵詞萃取及挑選出產品特徵詞，最後以各品牌產品的特徵詞或片語為基礎製作對應表並進行對應分析。本實驗利用二維定位圖呈現類別資料之相對位置，協助決策者能夠在一張框架地圖上識別出市場關鍵層面、消費者需求、產業間之競爭態勢及脈動……等要素，以提昇企業之競爭優勢。研究的結果發現，在咖啡品牌部份，實驗的咖啡品牌被分為四個群組，其中兩個群組為相互競爭，而在產品特徵詞部份，高密度聚集的特徵詞被分為二個族群，其一為將咖啡視為“生活品質附屬品”的品牌個性族群，另一族群為“愛玩、夜不眠”的年輕夜貓族群。

關鍵字

消費者自組媒體(CGM)；產品評價；特徵詞萃取；對應分析；定位圖

並列摘要

The conducive environment provided by the Internet Website has eventually transformed it into centralized source for assembling consumer’s feedback and comments. As increasing number of consumers are utilizing the Internet Website to express their personal views about various products, as such, these views or opinions are invaluable to business organizations. From the perspective of such business organizations, these information will review the preferred products in the market as well as consumers’ usage experience. For the purpose of assisting business organizations to captivate consumer’s demand and preference trend, this experiment will enable us to excavate or extract product comments from the various websites as the experimental infrastructure. Via surfing and downloading such information on these websites, we will then perform Chinese Word Segmentation, Characteristic-Word Extraction and select the characteristic word of the product from the extracted web context. Lastly, we would based on the Characteristic-Word of various brands or terms as the basis to construct the Co-occurrence data and consequently enable us to perform Corresponding Analysis. In this experiment, we had utilized the Two-dimensional Positioning Map to present similar types of information in opposite positions. This, will allow business organizations view the following on a single mapping framework: distinguishing markets based on critical factors, consumer’s demand, competitor’s product trend e.t.c.. Therefore, will increase the competitiveness of business organizations. The end result derived from the experiment performed on a particular brand of coffee can be classified into 4 groups. Of which, 2 of the 4 Groups, are very competitive in terms of the product’s characteristic word. As highly intensive characteristic word can be divided 2 groups. Firstly, coffee can be regarded as a “supplementary product in our daily necessities”. Secondly, coffee can also be considered “essential product to those young and playful night party goers”.

並列關鍵字

Consumer-Generated Media(CGM)；Opinions；Characteristic-Word Extraction；Correspondence Analysis；Positioning map

參考文獻

1. Blood, R., "The Weblog Handbook: Practical Advice on Creating and

Maintaining Your Blog," Cambridge, MA: Perseus Publishing, 2002.

correspondence analysis and related methods," Computational Statistics

"New Riders Publishing, Indianapolis, Indiana, USA, 1996.

6. Chang, Y. C., Chen, S. M., and Liau,C. J., "Multilabel text categorization

國際替代計量

網站部落格之顧客口碑評論分析研究-以台灣咖啡飲料市場為例

主題瀏覽