許又壬

Yu-JenHsu

碩士論文（2016）

Facebook數位足跡為基之消費者購買決策型態預測方法發展

Development of a Consumer Decision-Making Style Prediction Method based on Digital Footprint Mining in Facebook

關鍵字 Keywords

消費者購買決策型態、Facebook、數位足跡、資料探勘、機器學習

Consumers’ Decision‐making Styles、Facebook、Digital Footprint、Data Mining

摘要

在現今企業競爭激烈的環境中，企業除了掌握技術研發、生產製造以及財務管理等能力之外，行銷策略擬定之能力已成為企業成敗的重要關鍵因素；而行銷策略之核心在於企業是否瞭解消費者之購買決策型態。過去企業為了瞭解消費者購買決策型態往往透過費時費力的大量問卷施測以及統計分析；但隨著網際網路發達與社群媒體的普及，越來越多客群會在網路社群媒體上留下互動行為、文字等資料(i.e., Digital Footprint)，這也意味著企業有著另一種不同的管道可以更客觀地暸解消費者購買決策型態。因此，如何有效地協助企業從網路社群媒體上客群所遺留下的大量數位足跡(Digital Footprint)中分析出有價值的行銷策略擬定資訊實為現今企業提昇市場競爭優勢的重要研究課題。

本研究目的在於發展一探勘Facebook使用者之數位足跡以預測消費者購買決策型態之方法，以協助企業快速且正確地掌握消費者的購買決策型態，進而降低行銷成本與提升顧客滿意度。針對上述目的，本研究主要研究項目包括: (i)Facebook數位足跡之消費者購買決策型態預測流程設計，(ii)Facebook數位足跡之消費者購買決策型態預測方法發展以及(iii)系統實作與驗證。

Abstract

In today's competitive business environment, The core of the marketing strategy is to understand the consumer profile. In the past, companies have to use the time-consuming questionnaires and statistical analysis in order to understand the consumer profile；But with the development of internet and social media, more and more consumer will leave interaction record, text and other data (ie, Digital Footprint) at the social media website, This also means that companies have a different way to understand the consumer profile more objectively. Thus, how to extract nd analyze the valuable information from the large digital footprints at Internet community media groups for assisting enterprises to develop marketing strategies to enhance the competitive advantage, is an important research topic.

This study develops a method for predicting consumers’ decision‐making styles by mining digital footprints in Facebook to help enterprises quickly and accurately grasp the consumer's decision-making style, thereby reducing marketing costs and improve customer satisfaction. In accordance with the above purposes, the main research tasks include: (i) designing a process of predicting consumers’ decision‐making styles by mining digital footprints in Facebook, (ii) developing the techniques involved in the designed process, and (iii) implementing a mechanism for predicting consumers’ decision‐making styles by mining digital footprints in Facebook.