利用網路特徵和文字資訊找出競爭者關係

Abstract

在如今變動快速且混亂的商業環境中，組織一定要能主動且快速地適應環境才有可能存續及發展。組織可以透過環境掃描來取得眾多的資訊。在執行環境掃描時，取得競爭者資訊可說是最重要的一環。但在蒐集競爭者資訊之前，找出競爭者是首要的工作。找出競爭者可以支援組織對抗險惡的商業競爭環境，或者規劃攻擊目標以及反應的策略。在此研究中，我們結合了網路特徵與文字資訊來作為經爭者辨識的預測變數。此外，本研究還提出了一個多回合的判斷模型，來達到改善競爭者辨識的目的。就效果而言，兩個方法在辨識競爭者關係上都表現的良好。

The business environment is increasingly volatile and turbulent. Organizations must adapt to their environments proactively and swiftly in order to survive and prosper. Environmental scanning can be adopted by organizations to discover varieties of information. Competitor information serves as an important part while conducting environmental scanning. Before gathering the competitor information, competitor mining is a key and necessary task for company that allows a company to shore up their defenses against probable competitive incursions and plan competitive attack and response strategies. In this research, we extend the network-based approach by incorporating text information as additional predictors and propose an iterative approach to improve the performance of competitor mining. The evaluation results suggest that the approach proposed by this study performs well in competitor mining problem.