網路輿論可以影響現實世界,舉凡預測政治聲量或是反應產品熱度,也 因此,如何以更有效率及準確的方式分析網路上的意見文本,是目前最重要 的文字探勘議題之一。
此類的研究通常被稱為情感分析或意見探勘。本研究除了分析情感分析 與意見探勘兩者之間的因果關係之外,也對網路輿論探勘進行了理論與方法 的系統性回顧,建立一套可供依循的實作流程,並建議未來研究應加入意見 指標以檢驗意見探勘的結果。

關鍵詞：意見探勘、情感分析、網路輿論、態度物件、意見極化

 Online public opinion can influence the real world, including predicting politics or reflecting product sales. Accordingly, it’s one of the most important problem of text mining at present how to do online opinion mining in a more efficient and accurate way. This type of study is often referred to as “sentiment analysis” or “opinion mining”. In addition to analyzing the causal relationship between this two concepts, this study also systematically reviews the theory and method of online public mining, and establishes a set of practical processes that can be followed. It is also suggested that the opinion indicators should be added to the process.

Keywords: opinion mining, sentiment analysis, online public opinion, attitude object, opinion polarization